UNCLASSIFIED

A D U MBE' . .

AD495905

NEW LIMITATION CHANGETO

Approved for public release, •distributionunlimited

........ .. ...... . . . . .... . : - .. . .. ..... . ... .. - i . ... . . . . . . . - ; : . .. .. . .. . . . . . . . . . . . . . . .

FROMDistribution: Further 'dissemination onlyas directed by Office of the Director of"Defense Research Engineering, WashingtonDC; Ot 1969 or higher DoD authority.

AUTHORITY

ODDRE memo, 2 Sep 1970, DDC-T [mr.rehbock] ltr, 9 Sep 1970

THIS PAGE IS UNCLASSIFIED

Reproduced FromBest Available Copy

W/-A,

Reproduced From

Clow Best Available Copy

(Data collected, prior. to July 1961) A 1 D C

areltie~ rn~ anll 57 M

OFFICE OF THE DIRECTOR OF DEFENSE RESEARCH AND ENGINEERINGOr/

WASHINGTON, D.C., 20301

.-. ... . .... .

PROJLCT IIi NDSIGIIT

Final Report

LL . .

OCTOBER 1969

1k (Data collected prior .to July 1967)i# 0 •'U~IA--7I • ...

This dowipmt my be fwtber di•tributed by a f with1 -0* Office of the Director of Defense Research and Engineering

Washington, D.C. 20301

I

:.-= BEST AVAILABLE COPY

_ ~Ii

SPEDING• PAE B1LA - POT flMD-

PREFACE

The Dep,.-rtnoent of Defense, like many other large orgattizat ionb,pursues a policy of continual introspective examination in order toidentify areas in which improvements can be made. Typically, these staff"analyses are meaningful and of interest only to selected managers withinthe Department. On occasion, howevcr, the topic is of such general in-terest to other branches or agencies of government, or to the public atlarge, that wider dissemination of the report on the analysis appearswarranted. This final report on Project HINDSIGHT is considered asfalling into the latter category; in fact, an earlier interim report onHINDSIGHT received wide distribution.

The reader must appreciate that this report covers only a smallsegment of a much larger area of interest in the dynamics of R&D manage-ment. The object of this preface is to emphasize that the Project HIND-SIGHT study had certain clearly recognized lImitations.

The approach taken in Project HINDSIGHT was essentially retrospec-tive. Twenty recent weapon-systems and major military equipments wereanalyzed by teams of technical specialists to identify applications ofscience and technology that were not utilized in predecessormilitarysystems designed to meet roughly the same requirements. The evolutionof the new technology represented in each system was traced back in timeto critical points called "research or exploratory development (RXD)Events." The RXD Event is the basic quantifying unit in the study, andis defined as the occurrence of a novel idea and the subsequent scien-tific and engineering activity in which the idea Is examined or tested.There could be one or two RXD Events, or an extended chain of them,culminating in a device or component found in a particular system.

The teams of specialists identified 710 unique RXD Events, conductedthe historical traces, and described and documented the related activi-ties in terms of the differential amount of knowledge that accounts inpart for the increased cost-effectiveness of the systems analyzed, com-pared with their predecessors. These 710 RXK Events represent only aportion of- the events that might have been identified by a more exhaus-tive anrlysis. To illustrate the gross limits of the study, ProjectHINDSIGHT concentrated only on the post World War II contributions ofscience and technology to the selected systems. Each study team was al-lowed about three months to complete its research on each system.

In treating the sciences, HINDSIGHT distinguished (I) the basicresearch done to solve a specific assigned problem from (2) the basicresearch done to expand the frontiers of scientific knowledge; thesewere categorized as directed and undirected basic research, respectively.It was found that RXD Events from the directed basic research categoryemerged in systems development approximately nine years following theirconception, while it took 20 or more years for some events from the

B iii

BEST AVAILABLE COPY

Urld i I k~ I I'd t i 011 y I I' J u In )IT 'I"V(i ''pmt. I I tlli. lh.. ii III , I-!., i

notd ii t Itiedt I. i 1 lly 11- 1 1 AIi lii (,,II r! Ill]1 I W i Jr I i-c 1;1 .Ii 1-! i 1' . J3

lii ,IiL) of, Ih flIvtih II pud I~s I- .'l vd tIl- ;ii -pl l -Ie Iw' I~i I I

terprcted 1-. tfft-.en1 ill(] M~y 'ucl(I ev 4VIUnoItiye j Udc(IIIenolt (A~..tIe

PI ~iIUS hI V b.1, i iqi, i LI l Wh jci ( .- I I j udiit'ii t O I .( h, I K I;, i.1 i qJim L lit ulnt.c

P roJ oc t HItINDS IG~ Wili'ha prov ided datI -.hnwinlj tiht , h)1i od Iy *eik.q,L the DoD -; reqii i rom gt-L-, I re being lull iI Iled by Ilit- reso I I -. (.)I L),.t,:ii~e-

s;uppor ted rt-.'irch anld de ye ~I open I p roq John. Dur ing OIL- I 9jt.- I !J 3 III IJU-S I GIAT Iu e f rame ., i n wl i chl L)G perCent of' thev i dent ifid l(XU LVuilloccur rcd, the DuD spent about $it0 bi Ilion iii support uI* -.Cci en. and(LechniIoloy. Als.o lur inqg t hi , peio~i d , i idu~tn.ry .and othcer 1Fe-de ro1 agen-Cii's Spent abouLt $6 b.,iIllion in the sajme a reas. ILt m igl tli b exputcld,then, th,-t about. 40 pe rcent of DoD-ut i I i zed -1cience arid technoloqy wou hicolle Frivii these-other sources. In tact., Project H INDSItUII Fouild that95 percent of DoD-utilized knowledge camie Fronti scieunti fi c and techliolog-heal activities supported either directly or indirectly by the Departmentof Defenve. The DoD will therefore continue to ask foi the neces'..vIryresources to fund and manage broad research and developmcrit: progran~s

A that are responsive to our short- and long-term needs for weapon systems.

Some results (,f Project HINDSIGHT suggested that the interaetion ofscientific and technological knowledge is stimulated by, and is most.

* productive for, weapon-system development: in a problem-oriented environ-mlent . The dynamics and causal relationships between the researcher inýthe basic sciences and the applied scientist or the technologist are not,of course, completely understood or defined. However, the 1HlIDSIGHT dataindicate that 61 percent: of the RXD Events identified had a specificsystems requirement as an objective and that over 85 percent of the tech-nological events occurred after a problem applications group originallydefined the problem to be solved. This finding seems to support the

* hypothesis that the rate of transfer and utilization of scientific andtechn icalI knowlIedge may well be i n creased when a spec if i cgoalI o r sy stewproblem provides the impetus.

The Department of Defense recognizes clearly that the HINDSIGHTproject has nct studied, in any comprehensive way, the influence of sci-en~ce on the development of technology. There are findings here whichdemonstrate that the basic research community clearly contributes tosystems development-for example, through the fundamental process of ed-uicat-ing those who become the inventors and users-but the study did ncL.1examine the role of basic science. The HINDSIGHT report and other sub-sequent analyses of the interaction between science and technology havenot: altered the strong ph~l~olop/ticci commitment of the DoD to an,,aggres-sive, high-quality program of research in the basic sciences. Such aprogram is critically imnportant to our future national security.

IV

BEST AVAILABLE COPY

Int %,uluuatry' Pr~j e(tl HI NDSIH mad1etl a '.1 qo ilicut t un tonr IisO intm I owarda1 bet tot, uttdv-t rn t~d 111 ul' the luctcý '111d tqtiviroWtll t ht~t t ftP)sicr fit- IFI CW

it)t I chit l nijy i it t wvaput -'o/y- I emt leveI oput-initi. It. duo, tofil reprv "t-1t t a

Ita Itl off I I )ta I ans - e r ; r al t Ft!r, i t i, (-)ll c Irtte t i fI ' l Ito flat i otti I v I I tr

to1 enhatICt' 010 ttt11angeltnelt Of resttarclt and do ve I optttn Iresouu tes

iV

BETAALBECP

I; Preceding Page Blank

ACKNOWLEDGMLNTS

The design and conduct of the experiment, the analy-,is and inter-pretation of data and the report's preparation were accomplished by theundersigned and are fully his responsibility. Opinions expressed or

implied are not necessarily shared by the Department of Defense or itsprincipal officials.

Credit for the accomplishnient of this study is gratefully sharedwith:

"r. Chalmers W. Shlorwin, Department of Commerce, formerlyDeputy Director of Defense Research and Engineering (Research and Tech-

nology), who recognized the potential of a critical analysis of weaponhistory to provide guidance for improved management of research and whoprovided executive control over Project HINDSIGHT.

The designated members of the Project Steering Group and rep-resentatives of the Assistant Secretaries (Research and Development) ofthe Military Departments:

Dr. Richard A. Weiss, Department of the ArmyDr. Robert 0. Burns, Department of the NavyCoZone' James E. Paschall, Department of the Air Force

who gave support and guidance and assisted in managing the field teamsthat collected most of the data.

The members of the field teams, particularly the teams'

leaders:

Dr. John H. Rluth, Department of the Navy ,

Mr. Lewis L. Gober, Department of the ArmyMr. S. John Sereda, Department of the Air ForceColonel Paul E. Cool, Department of the Air ForceiMr. Joseph Dubin, Oepartment of the ArmyMr. Ralph Hightower, Department of the NavV "Mr. Raymond Brachman, Department of the ArmyMr. Wallace E. Hicks, Department of the NavyMrs. Barbara R. Patterson, Jonker Corporation

who, in addition to their- normal duties, diligently undertook the diffi-

cult task of tracing people and ideas through 25 years of history.

The many hundreds of-respondents who took time to search oldfiles and their memories for the desired information.

vii

• -"i

BEST AVAILABLE COPY

App. iX ad t nfl on *iq s texr-.,~ nfirt innf i nivwdt patt if. 11.1i Iq

who di vo I~u t' edn n t fit, p lift i ditudh' 1-(J d~ ill I i II tIe)11w I n

datai and c0Pl iId~rd)ib o qu id~nct fi' nt In- des i Uri o I Iw it- n' v ie imnt' t Pro jer t

Raymond S . I smMcolonIB U.S., ArmyDirector-, Project HINDSIGHT

vii i

CON II N'TS

Ackirwedqmenpalei

Execit ivt' Su mim-- ------ - - ------ - -y-- - -------Obj ecct ive.s anld St ratt.'(y --- - - - - - - - - - - - - - - - -

Me LhodO I Ogy --------------------------------------- - ** --- IV

HINDSIGHT Findings --------------------- - xvFindin rgs Pe rt inenit to St rotegies Used - ---------------------- Xxi

2. Objectives, Strategy and Methodology-------------------------I. Objectives - - - - - - - - -I- - - - - - - - - - - - -- 7

*2.2 Strategy and Methodology-----7-.-----------------------7

* 2.2.1 Fi rst Strategy-Signif icance ofIRecent Advances-------------------- I------------ 9

2.2.2 Second Strategy-Proportion of New TechnologyFinanced by DoD --------------------------------- 1I1

*. 2.2.3. Third Strategy-Factors i 'n the High Usage of.. AResearch Results--------------------------------13

* 22. Fort. Sraeg-Value-Cost Index---------------- 152.2.5 Definition of Terms------------------- ------------ 15

*3. The Systems Studied ------------------ --------- I----------------7413.1 HOUND DOG ----------------- ----------------------------- 17

3.2 BULLPUP ---------- ---------------------------------------- 183.3 POLARIS A-i ------------------ ---------------------------- 183.4 MINUTEMAN I and 11.--------------------------------------193.5 SERGEANT -- - - - - - - - - - - - -.-- - - - - - - -3.6 LANCE -------------- 203.7 Torpedoes hark 46 mod 0 and Miark 46 Mod I ---------------- 20

*3.8 105mm Howitzer M-102 ------------------- 21V3.9 'FADAC -- - - - - - - - - - - --.-- - -7- - - - - 21

3.10 AN/sps-48 Radar------------------------------------------21

3.1 MnesMak .6ad5-------------------- ------------- 223.1 Sarlgh Scp------------------------------- I-------- 2 I

3.13 Strategic Transport Aircraft C-141A ---------------------- 223.14 Navigation Satellite ------------------------------------- 233.15 -Nuclear Warheads M-61. andM-63 --------------------------- 23'3.16 152mrp HEAT-MP Cartridge XM-409 --------------------------- 214

4. Principal Findings- -- ---- ----------- 7--------------- --- -- -- -- --- 254.1 First Strategy-------------------------------------------274.2 Second Strategy ------- -------------- I---------------------334. 3 Third St ratcegy ------ ------------------------------------ 374.4 Fourth Strategy ------------------------- -------------- 5h. 5 Spvcific Challenges Regarding Do aaecn-----A, 1'Ii tiConal Findings-----------------------------------------

Qudlit For the D~--------------------------- 6jFMsniivrt io------ vct.------------------------------- 63

F-r~~r F v '..................................Y

CONI I NI S (toIi not)*

5. Addli t niti 1i F tli n g.. (ctfliuI i uS b i4 Idtea Protec t ion- -- ........ 6F

1-. Part icipatinq Orq-Lli i it -----o ------- ................- 695.6 Co•itribut iot. Scijtnt i tic antli lt'erlinical Pr,,mnnel -- 705 .7 Educa t ion--------------------------------------------- 71";.8 M; IItalry S e rVv ie. --- 7- - -.......- --............ ...... 785.9 Mobi I i ty--..........--- -I -- 825.10 'M arital Sta ut s ..------ 9----------------- -- I... ... ... ..-- 91)"5.11 Continued Postgraduate Education-------------------- - 51

6. 5.12 Sunmmary and Discussion ------------------- ------------ 926. Research Environimtnt fur High Payoff ------------------------ 95

6.1 RDTE Environments- ---------------------------- 7-- 956.2 Research in Science ------------------------- I .---------- 966.3 Research in Technology ----------------------------- _100

7. Requisite Level of Investment in Research----------- -------- 1077.1 Strategy and Funding --------- ----------------------- .107*7.2 Measuring Return on investment -- ---------- ----------- -1097.3 Program Balance--------------------------------------- 1157.4 Adequacy of DoD Research Funding ---------------------- 117

7.5 Economies of a Forward-Looking Research Strategy------ 1197.6 The Value of Marked Changes In Research

Funding Levels---------------------------- . ---------- 123'A 7.7 Cost-Effectiveness Controls in Research.4 Management------------------------- ----------------- 125

' -7.8 Relevance of HINDSIGHT Findings to Investmentin Research ---------------------------------------- - 126

8. Methodological Validity ----.--.--------------------------- 1278.1 Difference in Investigators' Abilities---------------- 1278.2 The Experiment's Repeatability ------------------------ 128

8.3 Basis for Invention Claims ---------------------------- 1288.4 ýPeriod Covered by Studies-- ---------------------- --- 1298.5 Investigation of Failures ---------- --------------- -- 1298.6 Bias of Investigators--------------------------- -- 1308.7 Distribution of Event Types------------------- ------- 131

References------------ ------ --------- --------------------------- 133

Appendixes:A. "Project HINDSIGHT-Expanded Studies of Research and-

t .Technology Which Have Been Utilized in WeaponSystems," memorandum from Harold Brown, 6 July 1965-- 135"B. Definition of a Research or Exploratory-DevelopmentEvent------------------------------- I ------------------ 139

C. RXD Event Description ---------------------------------- 143• ~~D. Suimmary of RXD Events-.... .... . ....---- 151D. Smmar ofRXD ven----------------------------------5

. E_ DoD Funding for RDT&E and Estimated RXD Expenditures--- 195

'S

5i.

I.St "att:ji~. i rran-.po I-t Ai IC1a Lt IC- IliA I A- 6,A: ANJ PIS.-~ 4 a- - -- -- - -- --------- 26

3. LANCE MIi s - -- - --- - - - - - -- -- -- - - -- -- -- - - 28.4. pi ne Mark ý6 and .1"7- - --------------- 285. MINUrEHAN I and I I IC 8 M ------------------- 306. BULLPUP Milssi iIV--------------------------------------------.37. 1IOUIJD0. DOC, Mi ss iIV ------------------- 7-------------------- 283. SERGEANT.Missile----------- ---------------------- 32,T9 POL AR IS A-1 kiss ilI------------------------------------------ 3410. Na3vigation Satellite (excluding bouster)-__------------------ 31411. Torpedo Mark 4i6------------------------------------------ 3512. In-House 001) Act~ -- - - - -- - - - -- - - - -- - - - - 3813. Technical Conifidence -Level ------------ ---------------------- 4III. Time Distribution .of Uti~lized RXD Events----------------.415. Age f Peirfiormer a~t Time of Contribution-------------816- Length orf Emp'loyment, l96-6 -------- I---------------- ---------- 8617. Payload/RangL Capability of C-l14l Araf---------11218. P Producti 1V ity of c-141 Aircraft7"------- --------------- --- ---- 11219: Relationship Between Improvement, Complexity and

Requirement for New Technology ------------------------- 120

TABLES

1. Project HINDSIGHT Weapon Systems ------------------------ 8I]. Classes of Research and Exploratory Development

Suggested in Project HINDSIGHT ------------------------ 12Ill. Undirected Research-----------------------------------7-- 14IV. Directed Research ------------------------------------- 14V. Utilization of Results of RXD Events---------------------*36VI. Modes of Idea Transfer ---------------------------------- 48ViI. Rate of RXD-Event Production, by Organizational Type--- 50Vill. Scientific and Technical Areas of Events:

Frequency of occurrence-------------------------------63IX. Distribution of Motivational Categories----------------- 64X. Motivational Links------------------------------------- 65XI. Funding Distribution Frequency-------------------------66XII. Incidence of Event Funding-------------------------------67XIII. Median Cost of Events----------------------------------- 67XIV. Distribution of Event Cost ------------------------------ 68XV. Idea Protection��--------------- -1 --------------------- 68XVI. Incidence of Technical Participation by Class of

Performing Organization--.------------------------------69XVII. Distribution of Rdsuim~s and Events by

Organization Class-------------------------------------70XVIII. Educational Level of Identified Contributors ------------- 71

xi

1T. TAILIS (coo. i nut-d)

XIX. Professional Mobility at the II.S./8.A. 'Level------- 14XX. Degree Distribut joats: Veteran-, arid Noniveterans------ 81XXIOf. 'bistributionr of Degree Fields by Educational Level-----84XXII. Interorganizational flobil'it -- 8XXIII. Respons IbiIi tyL-Autbiori ty Level--------------88XXIV. Unweighted Distribution: Degree vs. Responsibility

Leel- ------------- ------------------ ---------- 9XXV. Weightfd Distribution: Degree vs., Responsibility

-- -- - -- - -- - -- - -- - - 89--- -XXVL- Resfrons~ibility/Authot~ity Level vs.' Number of

XX~. Previous Employers------ --------------------------- 90ýV I. Postgraduate Educatlo--------7----------------------- 91

XXVIII. Typical Increases-in Ratios of Operational-Effectiveness improvement Enabled by Technology-------- 114

itXXIX. Relative Funding of Sciences'and Technologies,FY 1966---------------- -------------------------- 11 5

XXX. DoD) Funding for R&D Activitie---------------119

IrI

TI

zil

Ixi:C1ii IVI SUIMMARlY

Projo~ct III NDS GI wasi CSLd bdI Ii k lie d *ily i e. Vi rect-. of D~ ef I Isc RI.-" e rcli and [fuitili'I ii~ i i ill'l~d~um of 6 July 1 965 i to lt:u M',,~ ijlinS%'c'e t jr ies (Re'ice reh and Dtuve I opneii t )of' thle Mi Ii to .ry Depa r ImunIi(Slee Appendi X A~

ra: ~ ObJecti~ves and S ti-aeiv

TheŽ def ined ObjeCL ives cif Project H fNDS I GIT we're:

(a) *to identify those inana'gement factors that. are importantin 35sSUring that research and technology programs will be productive andthat p ro r am results Will be uti lized; a n~d

te ()Td mea s ure the overall increase 'in cost-effect~iveness intecurrent generation 'Of weapon systems compared to that .of' their

predecessors (when' such can 'be identified). that 2i s assignable 'to anypart of the D6D's' investment in research in science arid techinology.

The strategy adopted to achieve these objectives. involved:

_I.Determining the extent to which new weapon systems areactually dependent upon the results of recent advances in scienice or

* ' technology for their attained increase i~n system effectiveness, decrease'in cost, or Increase in. cost-effectiveness as ýcompared :to a predecessorsystem.

(2), Determining the proportion of any new technology, requiredfo'r attaining system characterlstlcs,-that was the result of DoD-financed

*research" in science or technology.

(3) Determining significant management and other environmentalfactors, as: seen by t'he research scientist-or. engineer, that appear to 'becommerisura'te wit Ih high utilization of research'results.

-(4) If the findings-of the 'f irs t .ttra tegy i nd ica te a s..ign ifi1- Icant reliance on new'.sc'ience or technology, devising a value-cost index(or set of''indices) which offers a 'quantitative measure of the return on'investmient in research.: in *terms'of the enhanced c-ost-7effect~iveness ofthe Weapon systeMs ma-de possible by the'purchase~d knowledge.

The Study efforts were organized 'to offer, in a comparetively briefperiod of time, 'the class of information t~hat could be-of prima~ry use tosenior managemen,,levels of the' Dob and the military services.. Theseobjectives have' been achieved. 'This report describes~the findingsT and,where appropriate',. suggests their implications. More detail'Ied i n forinia-tion6 which *although it might have some polity implications- is of

xiii

q ic J.r iuterest i to t~he lowel r nianagi'mlen t edt'IVO1 on.withii ii the ..C. it'll t i l'it-"and 1.echlilo loical tconti if i Iit--i vi ilth dt'v opd l' r

""lh'l study perfstllilt'l cConsisted of liht folhlowing:

Ad hoc t eamis of tai Iilto ry .111d civil Ianj i n-os I ( OUStDoti) pttrronmte I;iII

CdnsulIive tand part icipating contract support by behaviuralOlnil mant.gemei at sc ient is Lt- fr'om Northwes tern hiiveris ity and the Sloan

SchOol M•Issachusetts Ilst itute of Technology;

Contract support by econuometrics scientists from the Instituttefor Defense Analys'es (IDA) and the Rand Corporation; and

Voluntary participating support-from five industrial-management

"sciet ists. t S

Oki TThe in-house teams were concerned with establishing the amount of

new science -or -technology embodied in modern weapon systems, ascertaining

the financial suipport for this science or technology, and defining gross

management patterns. The 'univers i ty and Industrial-management scientists

are concerned with the more subtle aspects of research management, in-_

ciuding idea flow, information sources, training, individual effectivenessand motivation. IDA and.RAND are attempting to devise-credible and use-

ful value-cost indices for research.

Methodology

'Each in-house'team focused on a single weapon system. The subsys-

tems, components, materials and enabling system concepts were examinedin detail for evidence of the application of knowledge resulting from

recent research In science or technology. Subsequent to the recogniti.on

of this evidence, "each team undertook an Investigation to ascerta:in thesource of the new knowledge and the circumstances under which the know]-

4.• edge was generated and passed to the eventual user.

- The basic element for quantificatlon within Project HINDSIGHT is* the Research or Exploratory.Development (RXD),Event. The RXD Event de-

fines a scientific or engineering activity during a relatively brief

7- period of time that includes the conception of an idea and the initial

demonstration of its feasibility. There may be one or two RXD Events,or an extended chain of them1 culminating in a device or component found

in the studied weapon system. The final fabrication of this component

* or device may or may not involve an Event, depending upon the state of

the technological art at the time of fabrication.

Many observations derived from the analysis of Project HINDSIGHTdata confirm widely held assumptions of successful research.managers,

and in these cases the primary contribution of the Project is to make

available quantifiable support for what previously had been informed

opinion. There were, however, some significant findings wherein the

A'

i°-

Lt.I I rcont s-JkidiCt SO'il' q Irvi I hyle Idt bi- I I v i's Til- u lol w rig~ par~j rdphL,'con tain o 'uinmia ry of' I hio f ind i 69L , basedl uponi ona I y-ý i ý ol, thIr cur retn 1y

a aiI a b. v dat.) ;aind soirit. ill ttl)' 0~tJ 1.01 at On f' tlic,u,te f rind iriy

It, is t'iiiphas i ýed that th is.'suyidn ido .o. crieiIcoit r i but i JirS to VXi ý, t i[n bud i es or. s(; ei ii tI ic, and technol ogicltd kniowI-,edge that were Ut Iitilid i1n the naiaIy zed m i I i try equ~i pm~ents. The st rungtdIependen .ce: of the*se c Iont ribut ioi, upon the' total base of -science andtechnology must be re'cognilzed. .The reader is5 cau t ioned thlat airy cunc lu-s ions regardi ng 1he 'Val ue of ,the total knowledge base cannot, be val idlydrawn f romi the f intlings pi-esented here.

HINDSIGHT Findings

(I -Vh. E!m n~ct) J rpcJie~vft.; tit vazutluw or tv leeiulogy

1Vt1W'1PL'uh, thO rertn; t of' hiohi4J woJ)' ditLe in, and wer'e cori tir.-cl t&ne rw ~vwian oulnt1 to -be~ J'iriy viimple.

For mos t of the, systems studi~ed,- on the order of. 1,00 -to 150 post-,194*5 contributions from the scientific and technological communitieswere uncovered.'. Practical- limitations in: time and resources, however,permitted only a smaller number to be eyamined in detail. In no casedid the study team feel that It had exhausted the possibilities for. iden-*tifyirig'additi~onal contr~ibutions. Studies. of the indiviTdual systems

*were termnriated when the study team and the Pro-ject Director were quitecertain that a representative sample of the contributing Events had beenInvestigated or a 3-month studyý period ýhdd~elapsed, whichever came la-

study, varied from 20 percent for the C-141 cargo aircraft to 75 percentfor the LANCE mliss ile system: and over 95 percent for the Mark 56. and 57Naval mines'.

()The nwnmber ofEvents that were easily: ietfievaidro

potionate 1b with theý relative increase in efctvesbtentestudied 8"Ystem and Jits predecessor.

[For exa Imple, the studies quantitatively demonstrate that more newtecnoog ws rqure t prgrss:fro the HOES JHNroke to, the

LANCE missile system than from MINUTEMAN I to MINUTEMAN I I

(3) So category'of' performing agency was found to be significantlymore efficient in the production of utilized research results than othercategories.

apple drsearch utinds o iethesed categorens. tnvThe istrbutin ofidet~fid RX Evets ha t came fromuivr-

ties, industry, or the in-house laboratories of the Military Departmentsagrees, 'to w it h in a very few percentage points,- with the allocation of

xv

Il hough tile p rof i talh le wftt ký I requen t ly wat,. c i !. i 4)1e v', ap~pIi edresearch, if, thtat thle Scilenti1st wLm a~totept ing, to rt. lve I VU a ' V .pet %VCifiLprobl em, it does not appear.1h1at sped $ Ic alppiic-al ifon pewr ie a .i'uIfii-c lent cr1iter ion for predictilng usage. Perhap', thet Bios t ituportaitt I actors.in establishing o high probability of util/Iiat ion) art! thle deyree! ofaalr~eness (a) on'the part of the sc tent ~st concerning who it t hr ctngi -

!AF nee ri (~it conMunity oeeds the knowledge and (b) on the part oft the intf-r-ested enginleers as tO Which Spec if IcSCient i 't is wor-ki ng oti the prob I emt.Accordingly, research m~anagers must be sensitivce to the, need for Crytimo-nication between the scientific and engineering communitic!s.

Four sets of scientif ic Events that assume great Importance bc-rauseof the number of systems that rely upon them are readily dist inguishable.

1k They include work that led to information and filter theory, inertialguidance components, rocket 'engines and propel lants, and tran!istor Lech-nology with its associated miniature passive electronic elementst. I neach case the preponderance of the-research effort was inde 'pendent' of anyspecific weapon-system development, but was found to have been guided byindividuals who were acutely aware of the detailed problems confrontingdesign engineers and of the limits imposed on achievable system perform-ance by available pertinent technology.

(5) A signiLficant nwnbdr-'of' importtan7t. 6C.Si,;tififc cozt-r-ibue-~iou ca-&fzran sozoces other thanzcontemporarily recognized exrpertp .

Investigation of a sampling of these Events disclosed a fairly con-mon pattern. The performers of the utilized scientific effort and theirpeers were mutual ly aware of the exploration each was undertaking, butthere was marked disagreement regarding the merit of the approach thateventually found use- in each case the successful performer broke fromthe peer group, moved awa'y, and found new funding. More-than anythingelse, these examples point out the necessity for ensuring competitionwithin bureaucracy and the- necessity- for maintaining a multiplicity offunding sour-ces for every scientific discipli~ne and area of technology.

When high payoff from scientific research at universities was iden-I

tif-ied, the scientif ic ýreseaarch was coupled closely to research in tech-nology and some-limited development- Typical examples are the tnertialguidance theory and component work at the Instrumentation Laboratory,MIT, and the propellant research at the Jet Propulsion Lahoratorý,Cali-fornia Institute of Technology.

Within this report, undirected basic research in bcience. as tic-fined in Table Ill (page 14), refers to tasks not clearly Identifiablea,, part of a larger program that was coordinated by a mission-orienivdagency. In general, they are those individual, comparatively small ei-fort', ondertaken by one or two investigators in the univcrsities,,. in

~.nenot-for-profit c rporat ion's and, to a much sinai Icr cxztvnt. itr otiv-n'

XVI

i ritfuNt IY. I IItVy 01V L Il Ia -I.it V I iItld by I t'1 'I- ta Id Iha t Ithey V w t- 1110t i V tC

at Ithe I e& Lt0rChrtil -ý I CVC I by ,C icaht i I i L LUI W. it y arid 1,II I1 ly irrqt i Iy,anid worv ,tippoi ted by thIe DoD U1101 IeQ r O-4 I t ioh of tlt-I t- I tlV,11 rc1. anid i y-

Ani [mportjhrt paiyofl'it oin 'c i ti~ itif i r. rrear-li a on liii I% ye sI i~.- obeen Ohilru tile~schr SLI(.Ot edI st ci orr I i st 1u jid :ng i nteers h~o iateor pt-ifo rmnedtile. IPp I Ft d r e seorchi. OfJ th lit ],,/95 indiv idual i der itifi Ied as I 1,V iI IigCOil t Ii Ahit tLI~ to an 'RXD Event i dvti ai I d ii nformnat i (rn i% (;urr roitl 1y *ivai iJlab

oi .1 ,oir wo 921 perce~nt hol d acidtenic degjrees in %~rCfeIt or unq I rv -

Polleorly (e.gy. , KeplIe r's ,Newtonl's and oylosý I aw'% alidMaXWVIll'S eqaL tiOnS) suggests that a major payoff of tindi rectud basI5ILresearch is through knowledge acq u red 'dILur-ill rnidery raduato? education.

During FY 1965, based onl a percentage breakout, of research ariddevelopmentL funds, approximate )y 26.percent of the nation's scientificand engineering talent was employed directly or indi~rectly by theDepai trient-of Defense. Th is f igure has remained fai rly constant. overthle past several years. As a maj~or tuser of these resources, tile. DoDshould have a selfish interest in.,ensuring the continued supplI of tot-,en t by guiding potential ability into scientifi'c disciplines. or techni-calI areas of marked milIitary relevance. The selective use of researchgrant~s and contracts is one way of, affording such guidance.

Complementing this knowledge accrued through formal. education isthe engineer's later recourse to codified or 'tabulated information, such*a egineering handbooks and te chnical references., The second pyf.o

basic scientiIF ic. research is iden tified as the. organi zation. of researchresults into a. format readi~ly available for general use. 'in this way,new concepts of everything from materials characteristics to designtechniques have been widely disseminated. Management criteria intendedto enhance..the useful ness, of research .resul ts should .include recogn it ionof thie importance of .uni.fying or-codi~fying~Informatior, and the require-ment that the objectives of supported research be so oriented. Periodicsurveys should' be made, and research. advances in par~tiicular subject aireasshould be c~onsol1idated and, publ ish~ed i-n a format designed for use byengineers.......

* (0): The: gr'eatest payoff : ii z terins of ideas *Zeachd70 to eiialircc.;w.oapo). sy~rfemrn has reouZ ted from researech ini techriotogy--ad thcp:, .h tttie, :"cazvrcf cient,ý .t0' or~vi~ncer zwa. iintLmately awaro of pz'obýCrni?i othrý 'ptCations. i.?fgiineer.

The transfer of: sc~ience to technology and technology to applicatior,has been found to rely heavily on personal contact between individuals(see Table VI, page 4~7). The communication link is of critical impor-tance, both fur advising the-scientific comimunity about real problemareas and for disseminating sc~ientific or technological knowledge to theeventual user, the applications engineer.

(vii

k.. ti I I vt et'i o' in pt Ift I 1h1 m it v l,,t.~ .1w I p".11fl '.y*.t-1 ifi. d it*~iý pr'i't.'o -.Ud I y not tilt.~ i i of ol inP two fit I hitut -,( i't-fit i L

100, 2100 Or 300 advarn. i". eachI of' wilith d alotle i.. aI.t ivel y ;li(fi.iiqul l4'es hIlfIdrvf- tir di vt'r'.e, UitvJI1Lt-i In t Olvi tic fille1 d .)fil ýid j tin' fl~d it

J un i fied opra'tional wetapn~n ',y'. t m. Tilt, tlarkat tvri, ic1 tl cni I,nI iv,ifitncJM S t be cacifu 11y "interfaced" wi th lvtfst of ot hc r ad.i vat-, t i'subs t III Ciat ell by Project IIINUS ICH41 data.i which~ tat'ncvra IIy %how taN Ia !.y'.-tents app- I i ca I tnuns ra t hie I I Ialtne c i elic i if ii.u o -. it v anda t (If I -toay for advaticed sys t vms

AnalIys is Of the timeai relationship of an RXD Event's occurrence toit%. utilization in 'the studied system reveals two characteristic patternsWhere thi• weapon-systiem developinefit progressed directly front a technutlog-izal base (exploratory development) to enginec'ring development, th~ere isa mtarked peaking-of Events about the. time development is inu iated, withabout 41 percent ffor-LANCE and 60 percent for MINUlEMAN 11 occurringafter development, had started. This situation appears to be incoampatiblewith current procurement policies 'that stress f-ixed-price develoipmelt.contracts '(with or without Incentive fee). Where -the weapon system pro-gressed from a 'technological base trmrough an advanced-development phaseand, then into engineer'ing development, a mtuch more unif orm di str ibutLion,devoid of significant-'peaks~,.is noted; examples are the Mark 56 and 57Naval mines.

Project HINDSIGHT did not-investigate system-management aspects ofthe weapon systems studied. It is understood, however, that in at leastthree cases-the BULIPUP air-to-surfacfe missile, the C-141l cargo aircraftand the -LANCE missile system-development: was undertaken after-assurancewas given that the requisite scientific and technoliogical knowledge wasin hand. The Event-distribution charts adequately demotastrate that theseassurances were not valid. A major reason for the lack of validity be-comes apparent in a study of the detailed Event descript ons. At-thetime an advanced weapon system Is proposed, the. design e~agineer formshis judgments on the basis of experilence or -the extrapolation of thatexperience-the more advanced the proposed system, the greater the de-pendence upon extrapolation. A requirement for research in te-lnchniogyarises when the extrapolation proves to have been overly-optimistic.

In view of the potential, economic advantages of fixed-price system-development contracts, the matter becomes one of determining how, inthe light of Project HIlNDSIGHT findings,.it is possible to ensure thatsuch a contractual situation can be meaningfully established without.

xviii

IliL' M.irk. !6 and 1.i/ Nalv~i I minc., devl o~'Idpet- in ll l t(1W rolllll~it 1 11.11

c t r i 'Fl I I y tAO I d Ibt' con l', dtrk LI ' tllV,lllClilt dtvcloi~lllellt (10.0 f- aI ~o y o .- ?,

tJ ai, , IIi i i fit , tit ,it rs Iltiw', he L lt r ral y prFjIe'% i t ill~ .w hic tucholf-U(J lol11.,1 iV t I a. il. k' I jII tl t di)'- t Iit I' 0 tIt l thartl L t i(l . i .TI11(1I 0 1 OI'rI'

I.il lhl v I i vl.'y sc I iIu IU v. T lit- ii i ve % t iCJ(lat orll( di S lc I o e( t lit. (IUr i nlqth v.ily development lyear%, the work was accoinup li~hed alt 231 lnt ia I Iy

tons" tanl I I tv,,.I o f L,' f for t. F i xCd ýSUII Of money were iiadc - v i .1 a)I c for-

do~ i~it Lots bldpc.. Product ion did not. !,tart unt il a prototype deulori-

Thee ealiplt;Suggest.La the. greater use of prototype-systCIm devel-

omIl'lnt inteadvanced-doveIopmenot category, to Provide focus and sputto sorc in) scie1ct and te~chnology,. call be a successful preludeý to

fixed-price prodIuctiOn contr.1acting nnd still ensure a ,ignificant in-*creaise in opeirati onal performance..

iditi*o. oapt~ be. r.Lc9taiancL.l b. nr.

Al ternati've ly, fixed-price development contracting might be miade-feasiblo by as~signing to the Government project manager a degree ofcontrol over a portion of the Mi~litary Department's technology money'(R&D categories 6.1,.6.2 anfd 6.3) in addition to the engineering-de~i/e-oppilent (6.4) money committed to the contractor. This would allow thcproject manager to influence the development of the necessary technologyfor his system, e~nsure <that the technology money was inl fact directedinto areas of established military importance, and by keeping the fundin the,6.1, 6.2 or 6o3 categories give visibility to expenditures'. Itis clear that.-the use of this technology money must remain responsiveto the control of the Government project manager to be par'celed. out. asneeded.. Obviously, great care must be taken to ensure that these fundsdon't become a "bail-out" resource to; cover nontechn~ical managementdeficiencies. Examples of the successful application of this approachwere found in the develop*ment, of the Mark 46 torpedo and the AN/SPS-4radar.

(1)2% )o)hav required a onsiderable~cvnount of n~ew.cPJ.I tkl h(2,1 to11 fnd moo.0t of' it.

The relat.ive ease with which RXD Events contributing to the capa-bilities of now weapon systems were identified, the number and diversityof those Events, and the recognition that the'sanfe level of the ýutilizedscience or technology almost never existed at the time of engineering

Xix

g.~

.4

d•t'Vt? llii iop l t •f aii o e lit ttit fily A iliti,11 pyredenitl- r',It* k. tA�h*ll .c11 ibiliti tJodt'nolils t o•l3•C irtrevotctably thalt a Coillsitteiable ,l fiifll iI Aitil y niew I lini l-

O.ty wai, q I tqirted Itr ani advalced we'pll)(n sý,ý. I om.

I t is est I ma tetd tlie MiIi ta iy Depairtmifents spent about $10 i iI ionldu, ring 1945-1_963)%ihetih n1mst o• Lite identii fled Ive ij tixt.urreit i-.tv Appl-ridix F). Foi- thte imw yearl s another e!h iniate Place% expendtditurelý fo•research in sc ince and technology by indui, try and other- ti, ftlis,sectors of the .economy alt .appro inii tv ly- $7 hi II ianti. Ithtinding thatthe acquisit in o f 85 percentI. of the uti I i zed new scIent i fic and tecthil•o -logical inforlmation was fi.nadnced by the Mi litary Depa4rtmieinits biecomiteý.significant in view of tie relative size of the twolamounts expended.

We may corcl ude f romn these two observatio'ns that a great deal ofnewý know ledge was requl red for. an advanced weapon systemt and t he De-partment of Defense could not rely on other sectors of tlid economy togenerate it. This conclusion is strengthenied by the finding that, ofthe remaining 17 Percent not directly funded by the DoD, 9 percent waspaid for by defense-oriented industries and was thus indirectly fundedby income from previous Defense contracts.

(J:" I flunoJefe' of' technotogi eat inJlonnation i.,; at ratoaZ hdighi-t!i'tl w lt:in the ,tfe-,,nse-oirented cwnniunity.

Approximately7300 separate organizations or corporations have beenidentified as having-contributed to the fund of new knowledge.' About 20of theme had significant responsibility as prime contractors in the devel-opment of one or two of the studied systems. Despite the great number ofparticipants, over 80 percent of the RXD Events (but an estimated 20 per-cent Within the set of systems studied) are known to have found applica-tion in-more than one weapon system. At the very least, therefore, itcan be claimed there: are strong- indications that information transferwithin the defense engineering community is at a usefully high level.

The Project HINDSIGHT study did not attempt explici'tly to identifyalternative technologies used in. the. several systems where one of thealternatives theoretically might have satisfied all users. The numberof cases involving such a situation was so small as to appear statisti-cally insignificant.

(I) The perfoimance of the in-house Dot) ltaboratories appeamr tohe eonsir,tent with the nwmerical strength of their professional pervonnel.

After the sources of the utilized new science and technology wereidentified, it was possible to examine time-dependent trends in terms ofthe relative productivity of the various sources. As far as new scienceis concerned, the available data are not statistically significant. Amarked trend, however', i!s observable with respect to technological activ-ity. Early in the 1945-1958 period, about 51 percent of the new technol-ogy was coming from the in-house laboratories of the Mil ;tary Depart-ments, with 42 perg'ent from industry and 7 percent from the universities.

xx

By I19',,. t lit. b~l Iari. L wa'. rcvvtr,,td , ',0i 011 1) ~ P I( '[ I (-(Pil t i I lyI F (.m1 titLOeDt I'tWnt Ilabora t ot ii' e 583 pIWICerit I I'011 i ndus1 trfy , and 6 percLent ,I u0m1 tilt:

WIIV 1,IL i w. (11Cilll,ri Iy Lil i ViL? C" i t y 011-11 t ' d'ti Cki'lI C I-% ',UlII Wi, It ICl itL '% d r L II

*Laborait ory Itof- L lec t ron i c,; or I [it, nii t runlieit a t il (I.Laborat Iory .1 M IlT)

Ou r i fig~ LII SWhoICle . file, thle 'sCieliuI' I l~C .1nd tenqi live r ing -Arurtyth ofh le non -Dot) cominui Ii t y e's Sen t i all Iy tuaddrup Ited . wii i I I! the s trei.riy th (if jhvwIin-hous.e Otfl'onse 1dbora t~oriles less t han doubleid, Thus , [tic' chariqc' iit,rela I1i ye output Jdppd rF -10 reflect only I hi're Iati Vin c rcwlsi iln -,ruiml1.1

*Findings Pt-rtinetit to Strategifis Used

In i ters of the four stra'tegies eniployfod to: attain the pr ima ry oh-jective of Prject HINDSIGHT, the relevt fid -sae Summarized hiere.

*-Strategy 1: Determine th~e extefit tor wh ich new weaponsystems are actual ly dependent upon the' results of

................ .!.................iernce or technbloigy for their

attained increase;in system effecti 'ven es s, decrease inIcost, or increase in cost-fetveness as compared toa predecessor system..

Findings:. Markedly improved weapon systems result from skillfullycombining a considerable number of 'scientific and technological advances.

*..There is a high positive correlation between the relativesophistication of the .predecessor and successor, systems--or the relativeI ncrease i n the ir ef fect iveness-.-and .the amount of -new science or newtechnology utilized in the successor.

. Strategy 2: Determine the proportion of any newtechnology, required for atta~ining system charac-teristics, that was the reut of [DoD-financedresearch in. science or technology,

Finding: More than 85 percent of the new sclence or technologyutilized in weapon systems was the result of.BoD-financed programs.

Strateg:3 Determine sigtuif icant management, andothr nvronenalfactors, as-seen by the researchI

scientist or engineer, that appear to, be commensuratewith high utili~zation of research results.

Finrdi ngs: -The utilization falctor appears insensitiVe..to classicaldifferences in organizational structure or prof~it motiva~tion appearingbetween U.S. industry, in-house DoD laboratories, and uni~versity-associ-ated science and technology centers. It may, however, be sensitive todifferences between these types of organization and the classic organi-zational structure of universities.

Xxi

III Ilb I~r

1 ittltI 4nI 'fVI itI oI I4.) it I le44lLc 1n4% quilct.4 I ly lit-04 I I., t-. of4 proitn di n4 pI'e.ntfilentimlcal tvxp I d44J ii fnIn it tilt- j vtllver.

fh ) C'O I nLe U'%' Un I f i ILI U I I-4 I i f I L. I li~t- Wil Cod 44 (a i F~

"t; ellnt IIC itI ) IU4111,1 t lul

-la py ratit of 4."vOUval4 ir It IetLld gy41 e edt1t

SOif i t~pL' Of it'ui ptli~nt hat, been il Plitr ii U I 1wy '44LLE-sUl, uIdirdIto qieneralt i Ilj Ut ilized knowledge.

Achievemen~t. of a high comnbine,] inventifiveneSS, or in-gentui ty and uti Ii~ation rate are. dependent~ upon the time and spacecitiexis Lence of f our p rimarey fac tors -the recognition of need, a suurce

of des n hefom ~5 neuae aetpool, capital resourcLes, and

an adequate Communication path to potential users.

Stra! 'jy 4: If the findings of -the first strategyI nd I te a significant reliance on new science orLechnoAogy, dev i so a val1ue-cost i 1ndexf (or sot of

idices) which offer a quantitative measure ~f thereturn on investment in research, in terms of the

enacdcost-effectiveness of the weapon systemsmade possible by the purchased knowlzdge. ' Thisobjective has not yet-been satisfied. Relevantfindings serve primarily to define the problemfurther.]

Findingst A number of.-factors Identified through the studies madethus far tend to refute'the possibility that any simple or linear rela -tionship exists between cost of research and value received. They in-clude the following:

- -f)There is. a high probability that any given scientific ortechnological advance, if used In one system, will he used In many other

fsys tems. Ithus, any cost-value index must e ither apportion the researchcost among all utilizing systems or 'attribute a portion of 'the totalvalue added- by ea~ch ut Jiizing system to 'the ident ifiled RXD Event.

(b) Improved weaponf systemst or end-item equipmients tend to bea synergistic rather than cumulative-~consequence of the several embodiedscientif-ic and technological advances. Thus the task of apportioningvalue added biecomes complex.

(c) The relative amount of new scientific or techlnologicalknowledge required to achieve greater effectiveness, lower cost or im- -

proved cost-effectiveness of a new system, as compared to that of a4 predecessor system, increases with the latter's technical complexity.

xxii

li. aflyt k'udkil,.pp , OXI'I 10L I OilI l !ilt'.1stiillI IiVdl~ik .,JItllL' ,I d (All'.. u:c

a\ no I i l l) I ve two maii~t It'r Iijiu,ý t lie lelIphla. ed

FirstL thV SCOPC Of [hi" 'ItUdy Lind its) rk-,ul taill I iidiiq'knn iou1-

cernedfonlIy with (lie acquli sitIion'i of knowledge _-sscnt i dl toJ the~ I (!(.11110-

lg CVlIdeve Iopluell I of* .1 W0,1poll sy~I em. (QuesIi ouiv cocerin~ oi h the ust, orv1 IU0 oL I 010 V'UIiu I ,Of f'OSOarchi in the bL'hdivcora 1, lif. fe fivirorumlentl hOr SOC i. dlSC i 'I1CuS were not Oddlressod.

Seconid, none of "the findings described above Should h)Ce it Ypruas a disavo~wal of the, valut? of very fundamuent~al re!search in scienc. Itis -hardly likely t1,1t the transiStor' could :have been invented by peoplewho sought a snai ler, more rugged electronic signal ampl ifier but Vicre*unversed in wive mechanics or the theory of electrons in) solids; that asearchi for new power sources would have led to the nucleus of the atomlin the absence of the work oif Curie, Fermi and others; or that radio ort~fIephone commun ications could have been invented without the researcheff~orts of Hertz, Maxwell, M~arconi and their fellows.

* The findings suggo'st only that: defense research is most likely toculminate in utilization when deliberate attempts are made to relate theresearch results to-the problems of the military-services.

-xxiii

A /J

"I INTRODUCTION

Pl o jett I INDS IGIl'I i, si i.. tor ic,1I 'utvy of the proces,, by which ir

"S ietil if .tnd tecmnoloqi Ca l bases. 1 ftldudies arb ft1ust1d ott: -I

(I O the peopl" who canl he identified as having played igni f-. t cant rolvr. in th•t, acquisition of this knowledge, the mana'ement envi ron-

moll menhts in which they worked, and the behavioral patteins they exhibit;

(2) analysis, of the utilized technology, time distributionsbetween a discovery or invention and its application, uniqueness in

A terms of areas of.,practical application, sources, interaction betweenthe recognition of requiremenits and the generation of new knowledge; and4!

(3) econometric investigations of research in science andtechnology, estimates of the value-cost ratio of this research, and de-

. velopment of new analytical tools to provide management measures of the"effectivenes; of research. -

Vto , .The time frame of primary interest in the study is the period from"1945 to the present"

- The history of .mil i-tary support of research ,in science and technol-ogy in the United States dates back almost.to the time of the organiza-tion of the military services. :Although the resources available for Icommitment to these activities undoubtedly fluctuated with national pros-perity, congressional or executive moods and specific exigencies of the

services, a number of factors combined to minimize inherent difficultiesin the managerent .of whatever research programs existed prior to 1940:

First, until the latter part of the 1930s, the lack of mill- I -

tary interest in recognized scientific opportunities limited the numberand scope of potential investigations.

Second, the research potential of the nation resided primarilyin its universities, and communication channels between the academicand military com•munities were pract ically -nonexistent.

Third, the total funds available, either in'terms relative tocurrent annual expenditures for research or in terms relative to contem-porary defense budgets, were so small that their management constituted

" no problem.

'Fourth, the contemporary practice of supporting research out.

of the development program for which the knowledge was desired affordeda level of coupling, or visibility, that minimized any requirement f."separat:! justification.

4•. -

A j.

f ,

Asi

ny S 1l , oot~ i I 'AOi J I I iii. I.'i'c (pjv 1 01111 iit I (". I 111',c t'ii. VA I ci'l' I Vt

C it~.1LIit1i It t t it kIqh I' 1 0', V . u p r I I Ii I I VIIV i II OtIC'l I it i Wii I(,Vt.[

11 l kl ' kt uii v1i. VS h's 11.) 1 Ud I I ~ i. i''. itt LiiilILi qtIi) Cat Still ( Ii.i O I'

* S t .Ito p01 tili I il 1 pp 111 1 d Y lol f it. -a ' w- op k Iiut I i I...t fC )I-' d i 1ii ? I i tIP I i.I'dy 1,advISO Ich i. I titjo Ilysh t'rin, 1ui 0 1 %upin rea I ncj V on't I v'. t, I I I IIII r 1ch coml I (I Wi'h II I rltoluct Y Conci'p I odi f nt'I Ik 1)rc Isup`(.ortediofI i un wtt I y fro se aif I u Iap

mii IL Litct' IIi V'Ied ai it rV S , v pr)ceI L1 illu I riMl aid usI lily nIt knowl V I n

betwee~In Lito savace wepo oyn L-m' whc I ri tur all Ii fb Ill ic rIq r n I

wdneap wepons., Thi p emvid d L ut lit? i creat-,i fi or a (it r-adily LIII icrain level10rof expen I courep for research. spotdSpF.1~ "101 AUii-Wi

Between~~ 194 ind 9 anetimacpote Worl bi ýlnwspntb th

MC1ilitar DtopIitVIItIC for researchn S n Cis ichiene tund vtecnloy II F'Yd 1966 rmefo mlore know. dg b Lon was l progrmeadvfoatisen davosv r. Ion bot crofeltive

ond exbsdiLutie. termrstearatcfh.ndtr a gont hepita

BroetweeNDSI45T was163'a estaihdin oredr to0 determiine was inal ytichly

Mi tr*Dp thentimportneo research in science and. technology. tor the6

ff th unqens o afes rqiementsr fa rsarc, ifany;

management factors that significantly influence the produc-ti v Ity o f research;.and

Sanswers to 'many of the othe~r questions that hlave -been, raised*by conscientious administrators.

*Specific questions have been as varied as the background and inter-ests of the individuals posing them. However, most of the questions-or,.more 'precis .e ly-, challienges to the continued support of !research-may befiLted i~nto one or more classesý. Some of the. primary orientations aresuggested below. Each is: presented in the'form of a hypothe'sisto affordguidance in te rms of -pertinence of information.'

*The Department of Defense's, requirements for informat ion inscience and technology can be sati'sfied to such an'extent by resdarchSupported by the Nation Ial Aeronautics and Space Adniilnistrat ion and theNational Science Foundation that significant' reductions can be made inthe app] i cable portions of the Defense budget.

2

-AA

* I tit- I I 10 1 y ii itjllIVi A ýI)1 I 0ii 11 o tId~ i C Will k i p ý t n

tjito t i~t~ dhj vt~ 11 i t a11i itIti m t i ld l *,tit It J l pu t' t 11.1 pi I iti Ill

v ~ ~ I ~ I Ot- I. v 1I* t ttcl l l L c~ U~#d i fo' lit'i i I) a Id Ut ptSit r Ii 1 Lt(. I til tf

- 5 hnultV)qy ,inl Wu dlr it) thv fi Ii y vlI tvu. I i ye tinun'. maikt- Luirp- r'.-t ivie fm,t -

ofelI t' i ye'. ~. adna Iy',.e. till'. edL~l ,epard ifely i dcii ii i oh I r'Vou aich vfIot

lbS sitni ii cant portioni uu f tIIhe Ut I t'us re uerclu piugr rdma tok b eotlIe I at i ft)-i ll iivr , aer poinut of Iliargi i~u re11 1t tirIl..

*Tite'.o r of st it-tit if ic research ill Lthe utiuIvi-r-'. I Liv. byt he Dclpai wtnt of Dofeti-,v a.d at curet fudt Ici v s , , Jfli i f i cant I yafrects thle aai labi I I ty of tile best faculty taetto tuac~h s tudei S.

The requirement for judgment 'cri teria, litipl ici t in cachi of th eshypotheses, suggests that anl anlalyi ofS0 the historical utse of research-generited knowledge cannot alone provide fully satibfactory proof -or

refutat ion. Project HINDSIGHT recognizes this situation anid adepts sorteaugmenting hypotheses, the test of which will offer further evidence foror1 against. 111he previous set, and also will provide'somte guidance to moan-agoesnet where the-p findings are Iless t~han sat'isfact Iory. TheL supplementalset concerns itself with management factors affecting the Luse of researchresults. The concept imp] ki~t I n thils approach i s -based onthe 'argumentthat., if the mnanag ement factors encouraging the use of research can beidentified, current management practices can~be comparatively analyzedand modified, where and if necessary. 'The general themes of the propo-sitions tested in the supemental hypotheses incilude the following:

criteria-The strategies used by individuals and groups in theselection of research tasks whose results were utilized.

Control-The nature of the control environment' within which-successful groups work; the degree of freedom of choice and action bythe various levels iis-u'2vis their superiors and the l~arge'r envilronment;

thedere an ntue f athriyover fiscal and technical matters atýthe various levels.-

Skill11s-The way In; which new scientific and technical skillsare developed; the role the skills play In accomnLplishing utili'zed re-jsearch; factors influencingtecs and timeo oigit ne field-

'.the use and effectiveness of retraining and continued graduate: studies.

Interface-,Relations between the successful~ly performing re-search group -anid organizations involved in other phases of the RDT&E(research, development, test and e valuation),cycle; the s-ignificance ofcommunicati ons lag and distort'ion; the value o If liaison individuals.

Idea flow-Conditions in the immnediate~ (group) and larger(organ!izatT~o ei7tronmen ts that influenceL the kind, number, source,coimmun ica t ion and disposition of ideas for new investigations.

3

4,

C Tt'., S i u'L il k'~it]J IO11(IUIy t'iiiup ty i b jii hi'l v t-1l i .11 , I IaIko, 111a k i riq tit

.Projot~c IIINDSIGfIT is. detscribedt in det01t ill -sc~tion 2, "Ohjj~rt'ci~vu.,SSt rategy anid Me't hodology.y W1' Loncc UZ I Idt. y ,;I h1C p rucudui ui t, ort, IIIIM1iJ i /d'

iii tihe folltowing paracgraphs.

Task I., RXI) [vents

Research i n science a nd technology is a, continuing act ivitLy of thc

e ~ Department of Def ense. If this act'ivity is of Value, it would be logi-cal *to expect that a generation of-weapon systems would evidence theconsequences of this resedrch-.that it would be based not'only'on thescientific and technical1'f ou nda L1.onl o[ p~eciediiig generaticins, but on newknowledge as well I. This assumptioni provides* the basis for, the hkstoricalstudies c omp rising Task t of Project HINDS IGHT.

The Task I stud~ies constitute t he sole basi!s of this; report. For-this first task, 20 modern weapon systems or minIlita ry end-iem equ p-

ment hav ben exmind bytias of experienced scientists and engi-

in eers for evidence of science or'tedhnology tha It was no .t available. 'orwas not utiliied in predeces'sor systems. .Ha'i ng recognizbd the app~lica-tion of this new'knowledge, the team undertook a deliberate historicaltrace~of the component, technique or idea. The objective of the trace

* was to identify the people, the place and the time associated with the

generation, of the knowledge and..with th~e contrribution of significantadditions by wh'ich the techno~logy reached: the' level mahifested in thestudied weapon system.

The p r i niry ,reason for mi ntiating the studies through teInvesti-gation of specific weapon systems is- to-forestall, any doubt concerningthe value.,of the scientific or technologicalI contribution. The focus isv.on ut ili zed. resulIt s of. i nves.tmen nt.i n resea rch.: P rojec t tINOS IGHT. is notjconcerned with management of.,weapon-s~ystem development except when les-sons may be drawn from the role of project managers in fostering new

scientif ic or technological knowledge.._-

Thernature and depth of information sought by the Task 1, study

teams are demonstrated in Appfendix C, the. form used forRXD Event de-Iscr ipt ions. At the completion of Task t, 7,10 RXD Events had been iden-tified, historical traces had been made, the related, act ivi ties had beendescribed in terms of the differential amount of knowledge that accounts

C in part for the increased cost-effectiveness of current weapon~systernsover that of their predecessors. It is estimated that this number rep-

OA resents between one-third and one-half of the RXD'Events that mnighthave been idenitified by a more exhaustive analysis.

'I4

task 11: Behavidoral StudieL'Y

I a'.k I I 'sttd ite-d .Ire IVlctised on thv: t'l.jhvio•lal c le !S:i i t f l"int iiv ildueal', who C re I. i bt Ied I the u I i !1. d ,knowIedge. 1anggeueee'ee1 ,eialyp'.Iare per forei ntl detail I d studies of a J111p1 inyj of org§ai a in. , I)-ad irdi--vi dtals identiiled in Task I. Areas. of iives.tigitation include--hIu .if(:not I iteei Ied to--idea tr.ansfer , ski II developmnlt, net ivat iti, ut il izedinforea l ion souL'rces , and ýthv ro•e ofl the cclical liaison eltti . A- cacchsitudy is comeepleted, the results will be reported, and the seric.s. of

*-" topic-orie.nted e•ssays will be coisolidated acnd published by the Offie

.,of le'h Director of Defense Rvsexarch and Engi neeriteg, probably nut untillate i, 1967. It is anticipaled that the behavioral studies will be

continued.

"Task III: Econometric Studies";"Task" Il addresses the problem of defining and implementing a tech-

"nique (or techniques) for the quantitative analysis of the value of in-vestemeent in scientific and technological research-its value, that is,in terms of enhanced cost-effectiveness for the resultant weapon systems.A limited effort to achieve a quantitative value differential-aqain, interms of predecessor weapon systems-has been undertaken by each of theTask I study teams. To the extent permitted by security restrictions,their findings are reported here,. Although a more rigorous approach isbeing sought by the Institute for Defense Analyses, a useful solution tothis very complex problem has not yet been found, Efforts are continu-

2 ing, and success will'be reported if and when experienced.

The rimainder of this report -addresses the hypotheses stated onpages 2 and 3 to the extent permitted-by the 'data collected through thestudies and investigation conducted under Task I.

-Because a numberof research procedures were invented for thisstudy or were adapted with-marked changes from procedures commonly usedin behavioral and other social science research efforts, attention isgiven in section 2 to methodological details of Project HINDSIGHT. This

study, descriptive rather than normative in character, concentrates on avery special class of activities-research efforts whose results are

:-known unequivocally to have been uti~lized. By focusing-on this special

class, Project HINDSIGHT accepts the risk of some potentially significanterrors of omission. The-most apparent of these are discussed and theirpossible consequences evaluated in section 8, "MethodologIcal Validity."

*Note: Tasks II and fll were discontinued in favor of other

related DoD studies.

45ta.

tIt~' ft ull j ,' V I ' t 11 In Y',k I' ti lit- t I l t -d fa1.1vii 111 df Ii I I I cdt:: It pitIith -IJt I i If I ;¾ . rIl

.;P Ill; St'ltlck' .iid tediiiolotjy. lIn I vvi y I t'al wiv1v, I Itli ', i' va pIi -I i.lpoIJt. At 10111i oil (Ii Voln t o th tileI -, bt-haviorail , ntwtural tind tt..tliciniot i ca-l'Ic I V eIIvl I s .JioL adtlLtla I t, t1o wa trailt I anit' v It'it' ilt ofI t.11 recollttt-lte itd, i(.to; rinldt'd iit,1fqEItiRVIit 01* tllitot disc ip11 itt"'. UnltilI I Lil herCOnip I t'd , 11o con I U!.ionI COan be reached (In whetlIhe r Ori not i inp Ii Clitiol'.ii0 I1 a I. ti, It) It o ildg, ieI U101t i it t lie Ptly S i Udl I % L tesl L it' % Ir app I i cabJ Ic v c rto'.'.Othier disciplines%.

it l N DS I G 1IT is, a tudy oft a dyn atni c socioloigical s it uatLion. It car'-notL be JS .ms uted L hatL obsu'rvatiu.1ot and f jnd ifm% pu-r I aiilknq to anly y ivcwl

t ime w iI I retna i n yaIi d . vv r eX tended pet- i odb . h -wr' fore , the i'~ccsii If'or a continuing study or a series or' rueptdted studies is Suggestvd.For Uliis reason awd because of the (Iynainiic sit~uat~ion, th i ruporIt inuti tbe con s idored to be of' an iliterim nature.

2. HI CTI VIS * S I'IATl 6 ANID Ml lllOI)OLOGY

i t-e 1)r i11,11 y illiec tI ivt-,,n i l PrivjO IIINMIGM 1611~ t il i-Alvd by tileDi) rvc it) tit Beft~ist- Retarch and mtiincer i nq (tDUR14 ) it, July !9f,!k (%ut!Append i x A) Wf. were 1,0 loow":

(.0 to idt'lnt ity. and f Fi ro iy teS a.ib sh Ifigunfetifo tr' r roseorch and tecn wity prgm wch

hav be -S'OL i'l tell wi tI I u it/at it I of reult

pt-oducedL by these programts; ý(b ito) mesure Lihe ojver-alt i skc-oa it) coi t-erffctIveless ill thle culrentlt.qenerat ion of weapons sys feils compared to thei rprvdeCLVSnOr' (whet 'such can be ident I r iied) which -i-ass ignab Ic to' any part of the total DOt) invcs lmctrtin research and Lechhoh-ogy.

The DDR&E's miemorandual concluded several preliminary invest igativestudies undertaken by the Deputy Director of Defense Research andEngineering (esearch and Technology) and accomplished during the pre-vious year, partly under contract with private industry] and partlyin-house-'. The pilot studies had demonstrated that at-least thc firstobjective was rea~so'nably achilevable and had provided Insight withregard to more promising procedures.

2.2 Strategy and Methodology

The fourfold strategy adopted for satisfying the objectives wasrelatively simple In concept and may be summrarized as follows;

(1) Determine the extent to which new weapon systems are ac-tually dependent upon the results of recent advance~s in science or tech-

A_ I nology for their attained increase in system effectiveness, decrea'sel incost, or incr-ease in cost-effectiveness as compared to a predecessorsys ten.

(2) Determine the proportion of any new technology, requiredfor attaining system characteristics, that was the result of DoD-financed research in science or -technology.

t ~ngeenmt Factors ffeating Recearch and Afxptoratory )ccommArthur 0. Little, Inc. (Springfield, Vilrginia- Clearinghouse for FederalScientific and Technical Iniformation, AD 618-321, April 1965).

ýA Tvial 5'tudy of the Reoearah and IExploratory,-Devetopnrent oriiginsof a Wt~apon r'yatem-BULLPUP (Washington, D.C.: Office of the Director ofDefenseResearch and Engineering, 15 December 1964).

7

Ta b Iel. PRO L Ci .111N )S IG 1T WEAPON SYSJ1MS -

P!'A i~nait ~i qr.... Typ~e ofý ýe~aj)o i i.systemii

HOUND D)OG Air-mto-Surface Missilo1.BULLPUP Air-to-Surface Missile

POLARIS A-i Submari ne-Launched BallIistic Missile.

MI.NUTEMAN I Intercontinental BAI listic MissilI

MINUTEMAN 11 Int er co ntin en tal I Ballisti,( Missil e

SERGEANT Tactical Ballistic Missilei

LANCE Tactical Ballistic Missile

Mark .46 Mod 0 Antisubmarine Torpedo

Mairk 46 Mod I Antisubmarine Torpedo

m-laa 1051mn Howitzer

FADAC Field Artillery Data Computer

N/P-48 Suvilnce and-Acquisition Radar

'Mark 56 Sea Mine

Mark 57 Sea Mine

Starlight Scope .,Passive Night Vision Device

C-141A Strategic Transport Aircraft

Navigation Satellite Low-Earth-Orbit'Satell ite

M-61 * Nucl ear Warhead, Genera I-Purpose. IBomb

M '63* Nuclear Warhead, Missile

XM-409. 152mm HEAT-MO, Cartridge,, Ammunition forSHILLELAGH Antitank System

INote *il--CDa ta onyfThe-s-etwi:ý-systems a-re -fiisuf f ic ient tow wa r-r an t th'eir -inclusion in all analyses.

I -IC tin (.i) I'l eriul s i iin t 0 1 Ili I i nilit Il t ,d othe; e-viuunent atý4 fact oI~~0 s, 1 uVll bly I hit- resl wa I c 11 i enl 1 i. I t vite iq rive i I hat appva r I10

he c:l n u1 i te wit it Iti qh it. tliIi /.i i oul - vseairc-h resit • II%

lL) ifthe I i fd i nqs o! the Ilit sti!tiati'gy indit ate 4inil icanit reiiliace oil iew seiýCi'CC Or teholog100y, devi Se at vdlue-C-'ts i Index(or set of indices) which offerr a quantitative measure of the ret urn onl

.livestl•ent in ii'search, in Iteitis olf thJe enhanced cost -effe(ct ivene•s ofthe weapon s.ystenmt made poSsi le by the purchased knowledJe.

t 2.2,11 First Strategy-Significancu of Recent Advances: The rind-ings based on the first strategy were a key to the entire study. Toallow making quantitative measures within the other strategies, it was"necessary to establ ish the geneial dependence of new weapon systems upon

new scienice or technology and, even more important, to identify thuspecific contributions of research that were used. The pilot studieshad derfined three oh the critical factors influencing the probability ofsuccess in this task and had strongly suggested a fourth.

The fiat- ,t'•ZthiJ', to' was the recogni tion that the contributors,of scientific or-technological knowledge, particularly at the more basic

stages, are infrequently aware of many eventual applications of that.knowledge. Thus, in order to ensure identification of.profitable re-search, it was necessary to beg n-the search with the evidence of appli- --

cation to a specific weapon system and then trace backward in time thesuccession of conrtributors to the knowledge bank. To this end, 13 weap-on systems were selected for analysis in addition to the seven providing

S-the bases for the pilot studi~es. The total list of weapons and weaponsystems is shown in Table I.

The criteria for selection ensured representative samples of asmany of the main categories of Defense weapons as possible; whereverpracticable, systems having generally comparable predecessors were chosen.-No attempt was made to select systems on the basis of suspected highutilization of new technology.

The second critical factor Identified in the pilot study was the-sensitivity of results to the total previous experience of the investi-gator. In retrospect, the problem is easily understoAd. The investiga-tor Is asked to analyze a system in order to identify the results' ofrecent research. The most practical and efficient basis for the jidg-

;- ments he must make is his own experience. A young scientist .or en•ineer,no matter how brilliant, generally lacks background. This critical fac-.tor and the fourth, 'suspected one were circumvented simultaneously. Thepilot studies had suggested that there was reason to believe a contractstudy group would find its-efforts impeded by such matters as proprietar.interests within the laboratories of the research performers, the levelof its knowledge concerning the management of in-house and contract re-search efforts of the DoD, and the relatively low degree of private

* industry's responsiveness to an associate as compared to that shown na• Government agency. The adopted solution to both of these problems was

n~itv I) l i kXe I-.t k'n S I L i k (Ri .t iLi a ic IV lu Ii'n 01 11 l a, L-i' I i to ry and'pai Li 'i I I i~ ci' (I ip'l I ior" u .y.I w ii , iiip

'leeL tit- In Lu i h lanced I A0t'1 %n' tut vI II () "cI(c t Ili iiiiIo I c i-

o!011 I~ep k hao. tl w , i II W ' ~ii ~ . I I I, I Y l I 3 t u ., ii I, ulno tal cd b y * t l it i i he tiV'idlit' It on *)k I lj vit I i)ioo -111i ) I ' lit-'udM.iii W I- Deip,1i lll 'y I (iu hth klk tift I l-W1Ic lit iii M r013011y .I ntIm d Ih

JPIJ ril vaI dt yandut'[Uti~~ oft I viiipl l .ilit uli [I(11 Li lot en

A.Int r po ns toll ilk hII)I ' W, R~ I i~ 11-o~- 10 '0 hi 0' ~ i 1-1i V w icr~ d11 Uit,) rI ni wt-1

ie to Le d th wo rk o I to lis l ' I I n-Ltd l lu'; IILi ali,, bt' I hoWI 2V .I Ilt I vIi 'tl i 1 dc1 t I Iitedc were ntrprti i'dCC. [inh t'is gr tI V I ori lit-' pI tiry u I ont Ytei.'lt dxc ept 11,1 b id I) 1,10ii O i' rilu i I( t ili 51(13 I ors.11 .A

thel 1DeLpillty ViRll the'sC1.1101h arid t Tchioliyy) ! th Dietori.I oil PojeIctH111. VIINDSIGHT .3n pincipal me.'Igibes% of' thltc0 sin i Ic MII, an tehOlogiclt4.I -'

tudeam tha hadO dvIloped,11 thW ytm hseapon Systems. ocptj 1lLi ff wretp rla-I

tivelysnIId were somreor ed ha if)qu as grhat moe of IiIa the retsearch ýtnlid8* eveopm encrion byt o themwsdn under~iil thies aognzne foc esn

Aadietcneuceof the meetiDRE Rsor ngd betweenoqy thd iec~r ProjectHIDGTpIDIGTuidpincipal n h prrmingberous o utize scientific andl technologi-a

upthe.ametns t oa holld, devlpdthsytaied% inrmtionb waon ssthos.XI w'erets.a

ThcI other I u studyingetw ssymtehs wL mad up ofc -cn ist ndengiiieers~~~from the inutiarra izain that par ticipathed as pim

contdractr ionstequveopente of the weapon s ystetw ems The anje tHIcDSIpaTd

difficvaties, asr desribied in thpnrofucsiona toechnica editort foeroedn

countered., It appears to, be reasonably wel~l cstabli!shed that this typeof study miust be made by DoD in-house people., either ad hoc teams orassignedl personnel; each-of these alternatives offers some obvious ad-vantages.

ofaThe third and J'~inal or-iiicaI ,factor was the operational definit ionIofadi screte scient iIF ic or technolog ical .advance. 1,The def in itioan had

to be used by a numbe r of- independently operating investigators withsome assurance that each, upon Identifying the same. advance, would de-

*velop a muiua~lly corresponding description. Because the advance Identi-fied would eventually serve as the basis .for management studics, It wasdesirable that the activity described be of sufficiently short durationthat a single manager or fiscal policy would he involved and the researchperformers would be readily identifiable.

10

I I it-e end'. I iItt RXII I Vvtill llxd'v abl I ~hi i Iivf tv e ,~I,. II.111 L1 ~ 11'~~I *I~ id t 114 1~ tanent l ie d (if at I i Vi y td.tin 1141

Whiti-h11 tilt' i w.) ýT. -,lbili. bd 1(o 111 hI i.1 v Ihinll iol1 u01 14".1 Ile 1,Ml)

Vt v~I dI I I i Li. I Io fit- ri v. 'm i Itr hum,m l nit vu ni I ulv Ily",I ill I l,.i't t. I lit i of dII t ~ldl liI kil 1. liliI- I-i i- 'l O t11111 0( i 1- 111. i y 'VeL Ift 1 tX l li~ I )Iti-

II Jt ,11. A lto rt dLit Ia I led dit- jI Itt i on cifi I hit IIXI I vy7ltt (I eveli in Appeni-tiix If whih Ie jl.o d LIM i. "it erl" Cac .1lk ti V i t ~It I fit- prce!" by hitick

tit~i. IL I~e ljIVIe I ed ill .1! IIX II I v ii . tI ran '. lI cr. roil I rI )III I hvIe 1 1 .1i lniOva I or1 to [1ie %ecoutil-and ',u) LUo I'% titi I/ot i fill

Each Ieavim bt'g(.II 1 h I t dy bly e.xdtfiniiini ,y*J I vin doIJutivi.~ittI iOlt, ,111;j

I y.-i nt sy-'. 0111 (10%ig ) Ifi ll tleto l ls of makjor Comonil)01t S and11 "Al½.111 1CiIjUPniL.,

1i1ld di 5CUiA5,Iingj W011lchi I0. Ilklatters, with ti1t- wt* p)lflI%Y'. LII CIL' vyliopiit-iet ore-l-g~i nvers to obtadin cltt to~e 1 the Pot "'5 lb Wl! ueOf eWW se VIJonce or tehnology.Lach Chnoit lienl becamte a trXCC' i niltia tor; ioto the tJ rea Le.t ti' t eui I

possiblb t Ilii current user or Ili ite was asked fto itivat ify his sourct-

doc.umien ,tor i idivi dua I.Then I lie souirce was jiltteyi ewtod for I lhe .'Jill?

1 florolat lon , anfd 41hV Prl)ti!s5 wtS reeae unt I I ti he ti % oS Lat I I race oftile' central concept termiiillated with thle discovery (if' the origInlaillr(s).Upon (lhe, locat.ion of the i nit ialIly respons hible person, file perioud "if I he

:_ -rII Event w~as defined, andý the requi red inforniat ion wa's col lei-ted andidispl.1yed JS sfiwn-i in Appendix C.

To min imize tereport ing loaicf on the invetgtrnsranvn -

writing qual'ity, and'provide a check on the relative completeness ofdtaIlI in each Event description, a single- technicial-aditing service w+ias

estI.ablIIsihed. Theý technical editor was also respons ible for convertinlgqiuanti fiable informtat ion to a punched-card format, which facil itateddata proacessing, -and for collecting biographical ;-5risums from the peopleidentifie'd as pa~rticipants In the RXD Events. (In the latter effort. thet techn icalI ed itor was supportLed 'and as s isted by membe r5 of t he s tudy teamls.

At the conclusion of, its investigation, each studyý team prepared asutminary report, which 'serves two, functions: First, it highlights thet teamns ob se rva t ion s rega rd ing mana gemen t, f is calI, eiv i ronme ntalI o r te ch -nological factors that appear to-be associated with the conduct or spport of the utilized research or the-transfer of research results to

-7. ~ application.' Second, it presents a crude measure of: thie consequence ofusing new scienice or technology in termls of increased cost-effectiventssof the studied system as compared to that of its predecessor. Thesereports, Whose security classification varies fromn open to Secret, have

A been given limited distribution within the DoD.

1122*.'Sc..2 old Strqte-Proportlon of N~ew Technology E Lind-v

thle DoD: The second strategy of Project HINDSIGHT was to dete~rmine theproporl.ion of the new technology on which new systems were based thitwas a result of DoD)-financed research in science and technology. Thle

L oaý information needed for the~ie analyses was collected In the course ofpreparing the RXD Event. descriptions. In most cases (fth, gen eral %ourct-of financial support for the research worker could, be readily detirrminot1c.

I4

tI

Table Ic L. CLASSI'S. or RLSI ARCH AND I XPLORAIL)RY I)LVI..01.MM i

SUCGGErLL IN MUR~LCT 1IIINISIUGfl

AMI Class

lives L qt Losi ~L and Appl i oil mfa thema t cs and

thooreticdl studies Concerning natural phenomiena (R).

Sc ience, Experiicilctal validation of theory and accumulation.of. data concerning natural phenomenta (R).

Combined theoretical and experimental studies of newor unexplored fields of natural phenomena (R).

Development of a new material necessary for theperformance of a function (XD).

*1_ _ _ Conception and/or demonstration of the cap~abil~ityto perform a specific elementary function, using

Technology new or untried concepts, principles, techniques,materials,-etc. (:XD).

Theoretical analysis* and,/or experimental measurement'of the characteristics or behavior of materials',.equipment, etc.; ~as required for design. (XD).

First demonstr'ation of the capabil~ity to perform a.specific elementary function, using,'establishedconcepts, principles, mater~ials,* etc... (XD design).

Engineering [Development of a new mianufacturing,*-fabrication ormaterials-procesn tehiue,(kD mfg.)

First, devel opment-of a, coniolete System component,equipment or major element of such equipmnent, usingestablished. concepts , principles, materials, etc. (AD).

Notes: R -re-searchXQ exploratory developmentAD advanced development

12

it In. 1,i t i 'it' Iy few caiie', a bott 9J.5, percent *the! work h.~d been t~harged Itf4_n i d -: jia RED~ or overheaed Ifciotnt .1nd rccortP. were Tici ;iva i I al; I v or .

wt-re i nade'quat e to delfine lund iiq iniput tO t he .i~cuunt ý . Where, thisýOCCuIr red, dnd ill the i nterest', of Conservat i sm, it we', ai h iirari I ly he- I dthiat the fund joy ye-s not the ut even though it migiht have beenl re-covered ende~r ASPR-XV I R&D ( I .e . , htit lrovi sion in thtA':"Pi. ,*flA liOquA11lion, yjoverninq independent research and development)

2.2.3 Third Stratugy--Factor%. in the High Usage of ResearchResults: -The third strategy was to determine management and other en-vi ronimentdl factor-, that appear to be uni~quely.,correlatable' to highutilization of research'results. Based 'upon the recogni tion during t hepilot studies that investigative efficiency was extremely sensitive tothe experience of the investigator, it was assumed that a similar situa-tion would obtain -in the study of management factors. Frthis reason-relevant experiencý! was sought adcontract assistance obtained fo wuniversities that have active management staffs in their science depar:--

r ~ment~s. -The university people were asked to draw upon their experience for

-hypotheses concerning the relations hip between the environment and theprobability that research results would be used. They were also asked

-to design experiments to test these hypotheses, us ing the Identified RXDEvents as test cases for the experiments.

By October 1965,:several hundred ftXD Events had been cataloged.Abot half o' them had taken plac- in a rather small number of labora-tori,ýs or other organizations, while the other half was spread over a)-most as many-organ izat ions as--thereewere Events. In the interests ofeconomy alone, it was obvious that'a sampling study would have to suffice.Event~s were categorized iy' their contributing organization, of which the

0_4 10 most prolific were:

Raytheon CompanyI ing-1temcoi-Vought, Ic'Rocketdyne Divis ion, Ncorth Ameeican Aviation, Inc.Lockheed-Georgia-CompanyPratt'& Whitney:Aircraft DivIslon, United Aircraft Corporati-onU.S. Army Missile Comniand,-Redstone Arsenal

U.S..Naval Ordnance Test StationU.S Nva OrnaceLaboary White Oak ~

Wright Air. Development CenterU. S. Army Aun It Ions Command . P icat Inny ýArsenalI

Because it was' sens'Itive to the sstm t-hat had been studied, theselection did not necessarily Indicate the relative cp~ality of the' or- Sganizations; but it did provide an efficient way of economically max~i-mizing the sample size for detailed management studies.

In November 1965, these 18 organizations were asked to participate

in Project HINDSIGHT by making a,ýailable a resident management scientist

13

Table 11l. UNDI RI GILl I'lISEARCHI

o1. Obevation of phonomend.

2. F~onnulation of hypothesis.

3. Design of experiment to test hypothesis.

4. Conduct of experiment.

5.. Analysis and i~nterpretation of resultswithin the scientist's. f rame of ref erence.

6. Report to the scientific commu n ity.

TAble IV. DIRECTED R ES EARCH

1. Statement o~f problemn.

2. Morphological survey of -avail1abl e aridapparently *relevant knowledge for .apos~sible so ition, ,or deli berate searchf or new knowledge leading to a pr~oposedsolution.

3., Design'of experiment to test Proposal.

4. Conduct of experiment.f.5. AnalIys is' and interpretation of results

within the f rame of reference of theproblem.

6. Report to the "source" of the problem.

14

t to HdndIt I n LI it) I V xe t nI) ti I I I t dt- i 4qneVI l y I hV ini i vvr-. i t i~ eoi .t0 rv-, Ii den I I.lt .. c iettl i t .1, wav, cun%' itivi cI th, rab Iv Il~i ajt IPWA~t fi Vth - F;l~ 1--i.t , tile indt i vi ttn't I' -ý now ttit of hi, off,1i /.It ia oil

ait It I its Peop It, ConU Id if-.IVt, .1 tit' Is i tit I-J) I,'.11HE I I 1 P 01 I. tidy I itilt . ¶.etondliiallelivil woulId he a. '.oa ,IIdt I hat t pt(1 1p it. t dly i 111 t t. rc, I' wv e P4 I ot tt. i eI C

[h i rd , bect: '~ a tl I lie des. i tpna Ited ma ad('011en I '.t. I ell t i .. I wooM Ili flot hei i n1tIr o-ducod ito tilt tt!St hypo thvee'. tilt I I .I alit wIhe t'xpe r intient %. wert conduc L. ellhi a', woo Id he talin imom.

4 . W By December 1 !)65~ one otit hI oin i ver'. itI i e%. had r-unduc ted h~i i ci:'r ~~cotrtCs ot inlstruct ion to fami I iawl'ill the Ics, i den LI landileilln '. i enti sts

with stIudy objectivyes and the propos.ed 11101 hodo logy . Dis .1 r i hut ionl oftest ins truments began Iin Janulary 1966-; tiht- col leou on sand ainalysis. of.Ia ta a re expelctIed t o con t i ite wel I I Ino 1961. Primary anll y'..is is ltei.finll respons. i hi I ity of! lthe suopport i nq u i ve r. it1 itlie i r report%. will

be col lectc'd, col lated , credited and p1.til i shd hy Itle Project HINDSIGHToft ice.

2.2.4 FourtLh Strategy-Value-CostIndex:- Thle fourth 'st~rategycallod for- defining a value-cost index, or- set. of indices, offering atquantitative measure Of Lthe return ort Investment in research. F rom Lthebeginning, this was recognized at the mlost. difficult task undertaken byProject HINDSIGHT. Because the econormeris of thle detailed studies wasforeign to most of thie scientists and eangineers identifying the RXD,Events, the conceptual cost-value studies were assigned to separate

f.groups. At tile time this report was prepared, however, significant prog-ress had not been made.

2.2.5 Definition of Terms: During the pilot studies it was dis-covered that the current DoD definitions of research (R&D category61)exploratory development (6.2) and advanced development (6.3) were mnade-quate for use in recognizing, classifying or administering research inscience or technology. A more useful classification scheme (see Table

4F 11I) was gradually developed during the pilo't studies and, on the initia-tion of Project H1INDSIGHT, was adopted without change. The successfulapplication of this classification system throughout the studies hasdemonstrated its utility.

Although adequate to provide a functional description of the severalclasses of research in science an d technology, this categorization did

I'not serve as a vehicle for identifying motivational factors. Further, itwas obvious almost from the initial collection of data that the problemof semantics could represent a formidable barrier in the description of

scintiicresearch activities. Solely to Mniiz the Imac of thisdifficulty, definitions were ado pted for two terms used in this report;(I) undirected (or disaisociated research and (2) directed (or respon-sive) research.

The definitions, shown In Tables III and IV, are arbitrary, and arerecognized as describing only the two ends of the spectrum. For purposesof Project HINDSIGHT, however, they were found operationally useful, andso are used as a frame of reference in the pertinent data analyses.

15

4T.

Preceding Page Blank

3. THE SYSTEMS STlI)II I

I L wab obvious that thle selection of weapon systu ms and iiiil itary10,qui pmnts for study under Project HINDSIGHT would iiifluence Ahe find-ings. The nature and cxtcnttof this influence, however, could not beassL:-sed without prejudging the eventual findings.

"For example, the greater part of tile Army's weapons inventory. corv-sists of relatively ordinary items such as rifles, howitzers,. trackedand wheeled vehicles, radios for short-range communications, and thelike. For most Army 'operations, these equipments are at least as irn-portant'as the more "glamorous" guided missiles and tactical nuclearwarheads. If it were a fact that, 'for the former class of weapons, thescientific and technological base came from the general civilian economyand that, for missiles and warheads, it.was the result of military-supported research', any undue attention to either class in the HINDSIGHTstudy would lead to distorted findings.

/

In a similar vein, if little or no new science and-technology weregoing into modern rifles, howitzers, mortars., trucks, etc.--a situationthat seemed rather unlikely but could not be arbitrarily ruled out be-fore the study began-.then any findings'based on studi~es of satellitesand advanced radars"would not be amenable to extrapolation or otherstatistical inference. To. ensure that the findings would bereasonablyvalid, HINDSIGHT had recourse to a sampling technique -for system selec-tion, not a truly random sample, but a controlled one to achieve anapparent balance of equipment types.

' A The weapon systems and'equipments that served as the basis for thisstudy (see Table I) are briefly described in the following pages.

3.1 HOUND DOG

HOUND DOG isa jet-.propelled air-to-ground missile with a. nuclearwarhead and a stellar-monitored, all-inertial guidance system. It is

. . air-launched-from a pylon beneath the wing of'ia B-52 aircraft. -The mis-"sile operates in either a high subsonic or a supersonic mode-and has anoperating range In excess of 350 nautical miles.

M.m Weighing approximately 12,000 pounds, the missile has delta plan-form:wings in a canard arrangement; the wing Is located well aft on thefuselage and the horizontal stabilizer ,Is placed at the nose, The fuse-lage, an ogival nose faired into a long cylindrical afterbody, containsthe guidance package, the warhead and all other subsystems, except thatthe power plant is suspended far aft in an engine nacelle beneath thefuselage.

": 'Development of HOUND DOG began in 1956 and was essentially completein 1959.

17

o .-.-

3.2 BULtLPUP

O BULLPUP i.s an air-to-ground missile, opticolly tracked .oid r.ldio--'Conw.1,11d-quided,, intendeud flur usv ga ISiýS t tactical trtj . asucht v, raiis."bridges, tanks, truck cunvoys, siO ps and grOntid wcapun eulplaceieri t s.

Aerodynamic control surfaces are on the nose Of the mIissi le (canard curL-figuration), and delIta wings provide pri mary aerodynamic lift. Thisuncmpl icated, relatively inexpensive Wel'dpoll ha.s a des igned maintenanice-free shelf life of 5 years and can be loaded on an aircrafi in minuteswith o) checkout or other ground support.

Different versions of theBULIPUP tmissile use either a solid-propell-ant or a prepackaged liquid-propellant rocket motor. The smaller"BULLPUP A can del iver a 250-pound conventionaf''warhead over a range upto approximately 6 miles froin the point of release by the launching air-craft. The larger BULLPUP B carries a 1000-pound conventional warheadover a range of some -9 mi les.

"In operation, the pilot of the launching aircraft monitors the mis-sile in flight by radio and controls Its path by keeping flares in themissile's tail lined up with the target.

Development of BULLPUP A was initiated in 1954 and led to a firstproduction run in early 1958. The first production contract on BULLPUPB was awarded in 1960.

3.3 POLARIS A-i.

POLARIS A-I was the first of the series of submarine-launched,inertially guided, intermediate-range ballistic missiles. The A-I wassmall in comparison to-contemporary missiles of equivalent range, could"be launched from a submerged submarine, and had high inherent reliabil-ity and relatively low maintenance requirements.

Its solid-propellant rocket motors were capable of del ivering anuclear warhead over ranges well in. excess of )000 nautical miles with

-an accuracy commensurate with the navigational precision of the launch-"4, ing submarine.

Despite the important Increase in combat potential offered to thefleet by the introduction of POLARIS A-i In 1960, the system's develop-

ment had exploited surprisingly little really new technology, especially

in view of-the relatively much greater amount that was used in the A-3version.. Undoubtedly this was partially a consequence of the A-I's verytight development schedule, which allowed no freedom for long-lead-timeefforts. Pre-19 4 5 torpedo-launch technology-was the chief basis for the"launch-tube mounting-and missile-ejection system; a portion of the mis-sile's suspension, or shock-mitigation system was developed from air-craft arresting-gear technology; the inertial guidance system, computertechnology and the nuclear warhead were no doubt state-of-the-artdevelopments.

18

f38

lih ope ' ddv oue~It-e of I ht~. POLAR IS AIV iIs jind i ca~ted bythle faC t t hat not urnt I I Apr i 1 1 95/ we.re dei' in' iide .i: eyt ingt I lit-Pr'incipdll technlicl1 tquidtic I i n'_ , yet a POL ARIS ,ijbiimiarilit. Was it '"e i t,

*June' 1 960 -a I'lull 5 year,- alledcl o F thI e it [I hil0 t .1rye t (alt I~c iv19

3. MI1NUTEMAN I andII

AlINUlE1MAN is a 3-stage, sol iL!-proIpeIldnt , all inert ial ly gu idced.I ntercon t I ien 'taI baI IIs t ic iiiissi I Ic. P roj cct HIIND)SIGHT examined h

- Icc hnboly I cadiung f I rst to MINUTEMAN I and thlenr~ tu .1 t s uccessor,MINUTEMAN 11I. This was done: in arat tomtni to 'gain so~me wide rs land ing ofhlow. much -new sc Lenc and technology :are- involved-in-the iiarkdugaig

-Of' FunCtional Ily~quite siniilar systeirs.

The MINUTEMAN missile is about 55 feet long, reachcs a veloc~ity ofover 15,000 miles-per hour- in flight, and has a range of over 5000 miles.Motors of thie 'first two stages are irt metal-cases, twhi le the third-s tage motor, 'is encased in fi~bcrg'las:.,

The iner t ial1 glu idance -systemi is a highly accurate arrangemient ofgyroscopes running on practically f i-ctionles~s bearings. ý A -digi~tal com-puter contains a .memoIr y sectilon wh Icdh-s9tores ýthe missile's flight pro-gram.- During flight, the computer serves to cnmalpare the achiedflgtmaointa win* the desieed P rog ram, cal'culate sg 1 ~n cor;4ect ons -so as to

mainainthedesmd trajectory, and send sgra to-the- nozzle steeringsystem to'keep the missile on course.-

.'A reentry systemn contains the fuzibng mechan'ism and the nuclearwarhead. - - - -

A steel-'lined concr~ete launch tube 80 feet deeplhouses the missile.Surrounding the launch tube, under seVeral feet of concrete, is a two-

J level equipment room that is reached through a hatchway with *a hydraulically operated cover. -Thi s room contains -systems for communications,launch control and monitoring, and electric-power conver'sion and dis-

triuton.Aditonal equipmen required for normal operation of theriuche is inst-alled in a launch-s-upport building a short distance away.

3.5 SIERGEANT-

This tactical Army' mi s s iIe ca n .ca rry either a-conventilona I high-explosive or a nuclear warhead up to a-, maximum range of approximately

*75 m~iles. The 35-foot~solid-propellant missile .is fired from a mobilelauncher and is inertially guided ?to the~target. Because.-of--its high.thrust-to6-weight ratio, SERGEANT requit-es no-booster.

For mobility' in the field; the'SERGEANT missile-Is transported insections on conventional Army trucks that have been--slightly modified toaccommodate the system.' The warhead-is carried in its-own van, and themotor, guidance and:stabilization-f-in sections are moved on a semitrailer.When a missile is to be fired, a suitable launching site is selected and

19

0 1:•i ~~~the p-rincipal vohtich: d( lre!.t atlt eu.lol t"pl ter•nit j iý,Silu-

a.-,enihbly, checkout and prepn ation for liiring ci be Ucopleted itn iF;t mat ter of minutes

IX'velopment ,,f SERGEANT beg.tii in 1951t, jnd ji' oditiuti Ltoiitiittced in1960. It wa, tIOe of tilt IirS yt iyb t) t) ip omliy a itiA du I nr -de.ign L01i--ept. When a ma1l1function is detected during checkout, thle fai letd compo-nent is simply unplugged .and a replaceltient asseembly inserted.

3.6 LANCE

"LANLE , aniother tactical Anny mii. ilu system, is being developed as

an art il lery-guided weapon to support the field army division. This3200-pound missile is intended to carry a nuclear, cOnventional high-"explosive or chemical warhead over a considerably shorter range than

"SERGEANT's 75 mi les.

The propulsion system: uses prepackaged storable hypergolic liquid-propellanL, which enables the same, instant readiness' provided by solid

fuel but also offers simplified variable boost and"thrust terminationfor improved accuracy. LANCE is guided .by a unique inertial referenceand automatic meteorological correction scheme. The missile is direc-tionially controlled by steering the rocket exhaust.

The LANCE missile is about 20 feet long, and its cylindrical body

is about 22 inches in diameter. It is 5tabilized by cruciform clippeddelta fins with an overall span of about 4 feet.

.A Among other components of the LANCE system are a self-propelledlauncher, convertible to a lightweight towed launcher, and a self-propelled missile transporter-loader. With this equipment., LANCE willhave cross-country imobility and will be deliverable by fixed- orrotary-wing aircraft and by parachute as well.

3.7 Torpedoes Mark-46 Mfd 0 and Mark 46 Mod .

These torpedoes are antisubmarine weapons designed to attack tar-gets at all operational depths., About 9 feet long ard slightly over afoot in diameter, both torpedoes can be delivered by either aircraft orsurface vessels. They have essentially the same self-contained guidance"system, which enables the torpedo to locate and attack targets withinits search volume with no external assistance. Warheads in both-cases

" are conventional high explosive.Primary technical differences in the torpedoes are found in their

w.i. engines and steering concepts. The Mod 0 uses a solid-propellant fuelto drive a hot-gas engine; the Mod I has a liquid-monopropellant cam-

r. engine propulsion system. Control in.the pitch and yaw planes for theMod 0 torpedo is achieved by tilting a shroud ring that encompasses theAft. counterrotating propellers. Roll control is providedby fin tabs. The

, Mod I uses movable fins for control In all planes.

20

7!

Al

DLevIlOpMCFnt olf tilt MOd 0 :wd', Utind rtaken ill M8 andli ~Irc Iii,t d(- I i.- er ie'; to the fI'lect welt! 11'de in( 1963. Btoc~luse it u.As !,,illy (if I lit plitveii

cipntt.so Itilt Mod 0, the Mod I dt'Ve 1opile n t W1 li 01ci a] ttd ill 1963j,W,1 C01111 I e t e d in ho 1 0 2OL yea 1".

3.8 10.5nm Howitzelr M-102

Tb is how itzier was, d es igned for the roe of' a conw-1t L i ona I I ightar i e~sc-y pi ee although mtanty of it* character-ibtIiC dre uticotIVent iotal!

Mos t art illery pieces, once eniplIaced i n f ir i n iposition, have a imi tedability to traverse in azImuth (typical ly on the order of 60 degrees)but the M1-102 can traverse a f'ull 360 degrees. In addition,- iit~iprovedsteels were~used in the construction of barrel and breech, which petrmitshigher internal pressur~e and thus *a cons iderably greater rangL than its105nim predecessors.

In) spite of these advances, the weight, of the entire howitzer waskept low enough to permit. transporting it by helicopter.

Ea rl1y models-of the 11-W0 were made available to the Army in 1962.

3.9 FADAC.

The f i'el d artillery data computer (FADAC) is ýa general-purposedigital computer that. has been designed and tailored to assist the

.;art ilIleryman ýin the, solut ion, of f ire~-control 'and~survey, problems.

Classified by the Army as a standard equipment in 1961, this com-puter can be. used with al.l classes of artillery and anumiber of theArmy's missiles, including LANCE.

FADAC is about 12 inches high, 27 inches wide and about 30 inchesdeep, and weighs approximately 190 pounds. Despite its small size, themagneti-c-disk primary memory unit has a capacity of 8192 words.

3.10 AN/Sps-48,1Radar.

This system represents an answer to the current fleet requirementfor a practical "3-D" radar capable of simultaneously performing multiple.functions, such.. as a.ir surveillance, control of aircraft intercept,' tar-get designation for weapon-control systems, and provision of informationfor use in evaluating the air battle. Earlier radars were limited to.providing information in only t.. , f te trecodntsaiuhelevation and range; the designation "3-0" indicates the added th~ree-dimensional capability of this niew radar..

* The antenna of the AN/SPS-48, is mechanically rotated in the hori-zontal plane to a cquire azimuth inf'ormation. A 9-pencil-beam pattern inthe vertical plane,,with electronic switchi~ng, provides elevation data.The radiating elements are Windowed sljots in the transverse wave guidesmaking up the planar array antenna. The array is end fed by a serpen-tine wave guide.

21

The rada r was dt"e i gi ted pr imairi ly for ins tal .a13t I oft of Ou i dt-d-wi s, i I e(ri qt es and ties t royter- it isý ~m~pot I e i a i tdoreove por t it ha -

t ion', in ia ta arisittij frot. OflI p's pi tchi and rollI.

The Navy's.-*velopmetit of, the SPS-48 began it thfe eai' ly 1960s. Aproduct ion 1mo.del was 'installed in the USS VOIIDEN dur ing thfe I I r%. t halIof 1965.

3. 11 Minves Mzirk 56 and 57

The Mark'56 I s an, a ircraf t- laid inoored' tminie in the 2000-pound class,designed for employinent against subiiiarlnes.ý The Mark 57 mine is i(ts-'submtarine-laid counterp ,art. Both use conventional high-explosive war-

Z_- -heads.

Conceptually, thes wepn re'not significantly different from

thei r predecessors-sof wVorld War It. Because of advanrced design details,.Inew, materials and 'recent te'chn'ologic~al d-evelopments, however, they offerconsiderably Increased lethality, 5sflsjtjvitý~ and resistance to counter-measures, the ability to operate at much greater depths,rincreased-shockresistance to airdrop, and much greater overall reliability-

The major differences in the two mines Ilie in their ma'rkedly-indi-

(vidual means of dejiloyment-the Mark 56 IFrom the winig of a hihprformance aircraft, the Ma rk 57 from a submarine's torpedo tube.

The current version of-the Mark 56 mine was re'leased to production,in 1963; p roduction of the Mark 57 started 3 years earliler, in 1960.-

3.1-2. Starlight Scope"z"-

The Army's Starlight Scope is a night-ývision device that can beeither hand-held or mounted on a tripod for use in observation and sur-veillance when It is too dark to see with the naked' eye.'- The instrument,which looks-much like a telescope or a rifle scope, can also be mountedon weapons to enable aimiIng at night.' -

-Thi1s passive device employs the principle-of image Intensificationto raiseý the, light l evel1. in the vlewed'scene Above 'the hua ey'slee

k of sensitivity. it was released to field use in I1965.

3.13 Strategic Transport Aircraft C-14AI -

Th -141, a 4-jet-engine aircraft with a crew of 8 men, was de-,signed to airlift all the military services' combat or support units and

-itnilitary supplies, delivering then by landing or-by parachute, as re-quired, It can carry 1514 soldier's with their -individual equipment or80 litter patients and 8 attendants. Alternatively,i it can carry Abouta 31-ton payload over 4000 'nautical miles at an average speed of 422-knots.

22

flht! i rc raft i I I I VetL Iolq, ha% .1 wing %pI i )l of appi oxIill.i ly111160 le't ad a maxiu1. he i glt Of al iI10 t 140 f:et . Its se rvi e ..ce i ii i s [

1- 1111,000 ) et. MdxilliUm ,•pec'd d, 11. ,l t 1i It I tude of' 25,000 feet i ab ,jiuut

"50.00 knots.

"The 1•rtter Coill tl.et authorizinfg the develolpment of the C- 141A wasdated April 1961, and the aircraft saw its first service in October1964. A unique aspect of its dtvelopiiciL was that the initial contract"required the aircraft to be certifiable by the Federal Aviatfion Agency:fo|i coimiie rcial use in add it ion to sat isfyi rig m I litary requi rumciicLs

3.14 Navigatioin Satellite

This artificial satel lite, launched into a low earth orbit, iaakesit .possible. to determine a ship's position accurately regardless of"weather or the anomalies of. low- and medium-frequency radio propagation.

By monfitori~ng the Doppler shift in the ultra-high-frequency signal

e .emitted-by the satellite, the navigator can determine the instant of the. atellite's closest approach to his location.and the slant range betweenthe two points. Using time-Ieference information in the radiated signaland published data on the satellite's orbital parameters, he can computehis position precisely. .

The concept of a navigation satellite appears to have been consid-"ered seriously in late 1957 or early 1958. In September 1963, a satel-lite was actually used in ship navigation. The first truly operationalnavigation satellite was placed in orbit on 6 December 1.963, providingthe U.S. Navy with the most reliable, precise all-weather global navi-gation system ever developed.

3.15 Nuclear Warheads M-61 and M-63

These nuclear warheads, now being introduced into the weaponsinventory, are advanced versions in comparison to their predecessors.They use nuclear material more efficiently, have improved structural andmechanical features, and possess greater adaptability to design con-straints imposed by the weapon-delivery system.

"Because of security restrictions, the warheads cannot be more com-pletely described here; those considerations also preclude subjecting Ithem to the same type of study as that adop--d for the other weapon sys-tems surveyed under this Project.

"The study of the M-61 and M-63 nuclear warheads was limited to a

sampling of RXD Events that the study team believed to be generallytypical of advances in the non-nuclear components of nuclear warheads.

The samples were small and contained less detail than those of the other

18 systems studied.

2I.2...23

"A'F

3.16 152mm HEAT-MPi Cartri d me I-409 :.

The Sleridan is a new darmored reconnaiYsance vehicle deveIoped bythe Army to .replace the post World War II light tank. Its main arnamuilt

• features a 152mm gun-launcher that firels either the highly aCCurtu±e anti- ttalnk SHILLELAGH .guided -missi le or a relatively unconventional -r}:Itr-.projectile. The complete artillery round of which that projectile 'is apart is the 152mtmnHEAT (high-explosive antitank) MP (multipurpose) car-tridge X)M- 409. The "MP" designation means that the animunit ion can defeatboth hard and soft targets.

The cartridge case, which is completely consumed within the gunV barrel when the round it fired, is a two-piece assembly, manufactured by

a felt-ing process from nltrocellulose and Kraft fibers with a resinbinder. The projectile cbnsn•ists of a one-piece forged-steel body which

is loaded with-approximately 6.5 pounds of Composition B. The high-explosive charge Is shaped so as to provide,-.on being fired, an extreme-ly intense focused jet :of energy; this represents the weapon's antitankkill mechanIsm. Fragmentation of the projectile's steel body providesthe capability for defeat of soft targets.

.-Ow-

24i

[P 7-

4,.-

- I I l

4. F'IUNC 1PALI FINHNNG

I it, qtLcrt 1,, kbj ck:t ve, li I rdt jie v%. ol 1in ot' I Ii NU)SILI WV ILit'%C i IWO inI SeC t oi 01 2 , anin ',rmIt add (I i tio(I'lI ý.pq C i IciL qut-' I iol., if, W1i~ Luth. I ' t Uu(Iy Wd'. dd~ljI (eSkl wet- I SC se I' I 11 110 the Hit Io(Iuct innl. 11hr f i ittci lit)o f L 1i i' PVo j et: in1 ter' V i th0 feea r I W921 lO (tivi VV% w,~qi' dili

I'li' study approach adopted Ini Pre joc H INI)S I (IT is ,UcII i d I ,it

the pk1 iýýth W11sen , tht! f i ld i nys 1'r Iuiqu'Iit Iy aI*C i tidi cat ivL rajther thonlconc.I s lyve. This acknowledgmenrt in no way weakens the %lyni ficdntc (Athe t mdiniis , .1but recognize's orily that the completx it y of the processes!LStudied is so great. as to preclude an exhau!stive under'~tandimig. Morie

-pcfillIiy., i t concedes that j grea t variety of' manlage-meli pa toern'have been 'shown,. by IaIt Ileast Lone examp I o, to be workablIe. Neve rthelesscthec HINDSIGHT dilta indicate thdt.',soilic pat terni';, by the frequec icy ofLheir'appearance, are moure likely to be associated with research effort%whose results are utilized.

Wýhenever pos si bledata are' -pr esen .t ed as t ime-depIe ndeIcnlt d ist~ribu-Lions, or as percentagps of a Itotal,.to afford an estimate of the rela-tive degree of confidence that may be assigned to the val idity, of the

*indication or finding.

in some casesi, interpretation of. findings was found to be ex tremielIysensitive, ei'ther to the prectsiin' with Which certain words are definedor to the manner in which those words arc used. This problem was en-countered early in the study, and operationally useful definitions wereinvented or adopted. As pointed out in sect~ion 2,. the most. sensitiveareas involved-

*the general classification scheme for RXD Events and

*the delineation of basic research i.n sci~ence undertaken tosatisfy the curiosity of the scientist, as distinguished from th~at-under-taken'for more pragmatic reasons.

Classifying the sources of new scientific or technological knowl-edge in terMs of organizational type-DoD laboratories, universities,Iindustries-presented no especial probl em, except that it failed toprovide insight into the motivation of the performers. To overcome thisweakness, an additional classification scheme was adopted which~ dividesall research into two categories.: One, referred to' in this report asundirected (or disassociated) research, is conceptually described inTable ill; and the other, te'rmed directed (or responsive) research, isdescribed in Table IV. These definitions, or categorizations, providi. .1frame of reference for pertinent analyses and discussions..

25

Fl ur" SI milt-gv Ti-spl Aiivv-li CI 141:A

. , - .I'I$. -

s (C-

3~F1

1940 n945 1950 1955 1960 1965

Figure 2. ANlsPs-48 Radar

-- Predecessor System3 dev eopinsystem (SP) ront ract d.te

1940 1945 1950 1955 1960 1965Yea r

26

- I

i 'I

l "

.1 fi '.1St raeqy:l 1t'intri tIll(-xteiit to which litw wVapon Id ir rt'V aCtIraI~ I y depudentier Upoir tihe

re',U It I. o 1 ' 1" f I'.U 1 Sr i c crr U F oV I

itechnology for tiheir itlIainued i i ic re. vorvi* ~~~y t emii II ec t iverws',, dec reawe il IPA ,rr

()r i tic rea,C i n co. -effect i VLnes ., a.1compa red to a3 predecessor !,yslcmur.

* ~~~~~4.-1 .1 F i rid ing: Ati-kc il: vtr.'.i .'q)I !c~I*'i.: '.:: tf J~"vof'

A very-early observation of the Project HINDSIGHT study was that.,alImost wi thout except ion , no s inglIe i dent if ied RXD EvcntL or combi nat ion

* f a very row Event-E is responsible for an appreciable portion of a totaladvanced capabi Ii ty. It is most unlikely that the observation would holIdtrue throughout a -study *of the f irst atomic-fission boitb, wherein a rel-a t ivelIy smiall numbe r of very significant Events. involving Einstein, Fermiand a fw o .1ther physi ci sts apparently account for that capabilIiLy . Theobservation, therefore, is generally limited to cases *in which the prede-cessor and successor systems were technologically similar but unequal incost-neffectivenes,. Even here, however, there are apparent exceptions,as in the case of the modern artIllery howitzer. This seenting di~screp-ancy is treated later in the discussion of findings 4.1.2 and 4.1.3.

The time distributions of the RXD Events, with regard to the datesof decision to undertake engineering development of the particularweapon systems in which the new science or technology was used, formilluminating patterns. Representative distribution curves are!-shown inFigures I through 11 and are discussed in subsequent findings.'i

At

a.. indin: -There ia a hiigh pcjitzive cori-clatior. betweti; tiu!k-01-idsaticoation of'thc rdcno and ac'ceevi. era:i-g hincy-can in thi effectiveneas-and thci amunuvt of new:c.'

1:.".? t'c'IUh ofjy uiliized in t.he Succ'eiunor.

The argument for this finding requires the appreciation 6f two back-ground factors: (1) the numbelr of RXD Events that would probably have

beern disclosed by a more exhaustive treatment and (2) a measure of therelative value of each Event.

vestigated in detail 'Estimates of thiý percentage of coverage rangefrom 20 percent for the C-141A aircraft -to 40 percent for the AN/SPS-48radar. 75 e.-cent for LANCE, and over 95 percent for the mark 56 and 57ni ines. The ,p-repd .is a function of many factors'. The percentages varypositively with the accessibil'ity of the RXD Event's performers and with

!Summ.-ary des-zi ipt ions of these RXD Events and some of the morerelevant data are given in Appendix D.

27

:71

-. I.•

t940- 1945 1950 1955 1961) 1965

- Figure 3. LANCE Missile

ys~e~,(lit money available)

-First pCduclion contract date*

1940 1945 1850 1955 1960 1965Year

MAPN : *lInludpe inifiaanfI speeficiation rhanges.

Figure 4. Mines Mark 56 and 57

28

I hV " .i 10 nl d li i ul. OflL V 0 It' h ludy IVaill and, iuve r L, ly, will) t 1h1 -oiu-p e .it LIy of Ltile icluitl lltil'll t ,t it'd. "IC th e extt c.,uit l idble, L llWtVi.ve ,

etrCh %iiipl. I , 1'l)rle's.,llt Jti I V t O ih overall IItype Of RX Evv tilt cullii-!.-•~ ~ ll.l i t,I Lh' pe il " C q~ pinlicl.

nat Jq in e tpe Vfe Iq i pmenl

St udyi ny a It.l a| ively sllla1l numbe r Of Even t! tha i were -equejai;l ya-elated in 1 siingle setjlentiltic discipline or area ot ICLhiol-ogy was!genarre Iy preferrd Lt pursuing imany tentatiivl i dentifl ied Events acrossa broad spec •rumua of technologies. Thus, there could be up Lo Five t iies"as many Events as have been counted. The distribution amtlong the several

4r.- classes of rese-irch and exploratory developnment, though, should reimainessentially unchanged.

The seond factor, justifying all contributing Events as equal inweight, can be supported by a simple illustration. The transistor isgenerally accepted as one of the most important scientific and tec hno-logical advances of the current period. Of what value would the tran-"sistor be to an ICBM airborne guidance computer if the capacitors were

W Ilarge, leaky, wax-impregnated devices? ... if the inductors and trans-formers used huge iron cores? .. ,.if the power supplies were lead storagebatteries or internal-combustion-engine-driven generators? .J.and if thechassis and mechanical support were fabricated of low-yield-strengthsteel?

Recognizing the part that each advance plays in this mutually sup-porting effort, one cart conclude pragmatically that all utilized Eventshave essentially equal value in a single weapon system. It is thereforemeaningful to compare systems simply in terms of the number of Eventsrequired to achieve their demonstrated performance.

S• "The simplest weapon~system examined was the M-102 105mm .hOwi-tzer.. This weapon shows an increase in effectiveness over that of its prede-

:-¥ cessor.. the M-2A1, in two significant parameters: To achieve practical.transportability by helicopter, it is much lighter In weight, and'it of-

A* fers an approximately 40-percent Increase in range. A significant part ofof the weight reduction, along with the ability to withstand the higherinternal pressure commensurate with the extended range, is attributableto a single RXD Event involving the production of better steel for the

* Event of welding aluminum. A few additional Events involving propellant

manufacture and barrel-erosion control provided for the increased range.*Thus, starting with a relatively uncomplicated system, a total of fiveto seven Events resulted in a markedly improved howitzer.

Conversely, for a-somewhat: comparable increase in effectivenessfrom MINUTEMAN I to MINUTEMAN If, some 50 technological advances wereessential (see, Figure5). Thus, the greater the sophistication of the"predecessor, the more new science or technology is irequired to achievea useful imlrovement in cost-effectiveness. Where the predecessor"systems were roughly comparable in sophistication (a situation that gen-erally existed for World War II free-fall bombs and magnetic-influencemines), the study's findirgs suggest that the relative increase in

29

Figure S. MINUTEJMAN ! ;ml 'llv

MMINTIAMAN 11

"INUTIIAN 1 riro vhiti

"w- I I""

rI I I I I

I

"1940 1945 1950 1955 1960 1965

Figure6. BULLPUP Missile

j--System develolel

1940 145 1950 1955 19"0 1965Y-t Yea r'

30

C tI eiLi vt , o f e h, kie ' Lr•decs sors vri- d iurectIy wirt lhlI le nulyb r t

Litp t il tod rts. E thr

Th I ampexample, inme 110 to nm Ev'rto woryt requidred to a nhideve ui to

•A.A•. crvidse by a factor of about 4t between itt! free-tall boun b aro d thev BtLLPUPr::,"! ai r- t o- s urf a ce m I . %iIe . It toJok over 100 RXD Events to ificre;ase thv,

foMark 56 dmsine'd n effetinesst-affctiv.or Hf 10 over that uf the nark io

V•. iiitine. In each case, pr-edecessor ard succt sor are compared only witli:::-..'.respect. to role-5 that both c:ould perform.:

SOther paired example-. that indicate the same trend Lan rle cited.

The sample size, in terms of number of systen,: studied, is inadequate to

Sprovide a useful quantitative predicror of the number of Events e.sen.tiaSfoDr a desired increase in cost-effectiveness. However, crude analy91 is of

.'the data suggests that such a predictor can be achieved. This matter is

atreated further in section 7.

Mapnly, the systems studied were of the post-19so5 period.

SProject HINDSIGHT adopted flexible criteria for the definition of

e Fuor recent technology. of the technology of interest wal used in acruder form in the predecessor system, the evolution of that technology

was traced back only as far as the time the predecessor system was de-veloped, generally after 1945. If it was not used in the earlier system,the evolution of the technologyrwas traced back to about 19i5. The cri-teria were relaxed somewhat for new science, accepting some RXD Eventsthat occurred as early as 1935. Despate this double standard, 91 percentof the 710 completed RXD Events are classified as the results of researchin technology,"the remaining 9 percent, as the results of research inscience.

h Further, less than 16 percent of the technologically oriented RXDEvents were traced to a past'f935 science base. The other 84 percent

rcame directly from the applicalion of nineteenth-century unified theory,were the results of empiricas research, or appeared as inventions not

•.k:"•: needing sc!enti fic explanation.

JUThe median Lost of a DaD-funded Event of a scientific nature (R)

was approximately $60,0O0; for a DaD-funded Event of a technologicalnature (XD), the median cost was about $45,000. In view of the differ-

A ences in cost and the percentage of RXD Events oriented toward scien~ce,one possible conclusion is that a budget for'research in mathematics andthe physical sciences should be about 10 percent of the size of one forresearch in weapons technology, where the classi~fications shown in TableI I for science and technology (R and XD) describe the total package. and

1W . where the sole objective is gaining knowledge for application to weapon

systems.

31.4

4*,

Figure 7. BOUND IX)( Missihl

A.1 V

f-Tf- r-1 11940 1945 j 509055 1960 1965

Year

32

- j:H

h11 k 11' t.~ I d. i Oil Cannfot beL .31)1 it'd1 al bi I IJ31 I lY dw. I t c- t 0f L ur r(eri I

Dt,)) f tie dki jIIL)CdlIi tillS- for' .11It led.t LWO S2J91ii ILO Idi .e'iis i,I , t hekOlicIu-,itlin i'., .ppl icabie tnily wl-erL ithe :wor-k vl foil-. aii divided( iindc

codnti %1 (1 lthe dot ii nit 1'0n% of ,labl I I . Thie u1110-A (Llrsory Vxaminiial ioinof tIi~ hel) l , Cur1 ren I rtSe.)d rc h. arid vxt, I or i t or-dv cmLItp 4j t.rt tJrL-

tLi re v.hore, hflat a (L ol1S jiLtit- ah I porI-t i on of i Is 14 -se c Ii ci e FIortst ar e.1 j

rJ.)Li I I y (cC bLC110 in; i Cd 1 d (1-.1 tIeI t 11,111 SCiVieI : I Fi C Cildi IL tel'. S-COl~ld , ileID-ýpe Lt'rfl lCoctr dot-, not lilloIu for vWry ncsirwork in th litv iF,behav i ra I , t~oc id andi other 'a jenct-s

.,2 Second St ra LL'y: Determine t u'he p roport ion of any newtechnology, requi red, for. attainingSystem characteristics., that was theresdil of UoD-financed resecarchm inI . _ science or technology.

4.2. 1 Finding: '/h';( I IoP Jirimtiroorl t/w majuW)iIty of' Joyraufl.v .rpf,111-t' cS i'~ 01Q tt'violflogyJ~ J'()?- b~capofl-aydter wprwiiel

Throuighout the 1945 -1,963 period covered by this study, there was.almost continuous change in research budgeting, accounting and procure-inen t practices. Records have been ret'ired or lost, and seemingly littleattention has been paid to historical accounting for expenditures on re-search in science or technology. As a consequence, no precise figuresexi~st 'to describe D6D or other investmerit .*ii these areas.

:A useful estimate can be *derived, however, by noting that the FY1966 Defense budget-for these categories of research is $1.'39 billion.Using the generally accepted understanding that this investment had beengrowing at a rate of 8 to 10.,percent per year,. a total investment of

.S7.5 to $10 billion for the yea'rs 1945-1963 is calculated. A similarcalculation suggests that non-DoD funding (including industry, statesupport of university research, and other Federal agencies' expend!-,ture-,) came to some $5-i billion for-the same period. These estimatesare given primarily to demonstrate the absolute and rel'ative orders ofmiagnitude of expenditures for research within the United States. (seeAppendix E.)

For the 710 RXD Events invuestigated 'in detail, the distribution offunding source .s was as follows:

* .Dol) direct funding-- -- ------------------851.Defense-oriented industry - ---- -- -- 9Commercially oriented industry- -- --------Other Government agencies -- -- -----------Universities (state or private funding) ->1

The 85 percent directly funded by the DoD ircludes work supported bythe in-house laboratories of the Military Departments and all DoD con-tract and granit-supported work with any external organization. It does

33

A

itgur' 9. VOLADIl? A-i MissoiIu

Svmt dvefqt~ lvilvod ractt dale

4A

F-I

1940 1945 1950 " 1955 1960 1965Yea r

Figure 10. Navigation Satellite (excluding booster)

- Systelli developmeontcont ract date

U .

z" i

1940 1945 1950 1955 1960 1965Year

34

A

not incluide work supportedc uridc r t lie ASPR-XV I R&D Prograin, afunding pro-cedure under Whicli lefL-tie cont ractor,. cdf recover some 5el r-generatedresearch expenses as anr overhead cost against: bOD corit ra1.; fnor doe itinclude work funded by i ndus try out of* prof it5 on DoD contract-, aridinitiated for the purpose of gaining future contracts.

inclusion of -the last-named categories suggests that, whereas theDOD only funided an estimiated 55 to *65 percent of all on-going researchin science and technology, 94. percent of.the work that led to resultsuseful in weapon system% was in the class direct~ly or ind~irectly, funded

*by the DoD.

The principal conclusions that could be drawn from this. observationare that DoD requirements are in fact unique or that a tremendous duoli-cation of effort was involved. To a certain extent the latter possibil-ity is refuted by the findings discus'sed in section A4.3 regardingmotivation for utilized re~search.

Preck vessor . - System developmientSsyeme (Mkc 44) eunt ract date

A1940 1945 1950 1955 1960 1965 -7 Year

Figure 11.. Torped o Mark 46

35

c L

44

1 4-

W4-' 00

U 0-

41 C,)

u.* a4 4-4J

:< > 4-)

11- 0 0 II)0 C4 -Lf ~ ~ 5- f.) 4JL

CD Cm.>< ZA

44-W

ot

4-h (0

Ll~~ 4.'.0

4-or

~It*

CP

0 r 0 1M 0 4

, u4 S- -- 4)(

'A- r- W 5-' V) S.S0 -) CJ 4-k = a

0)4-)tz

0)36

4.2 2 1'i nd i oll til- i iri t,:I~.;J UjA ; : cc, :~.n'r

Wit ta thr- Staiiardt ie.dcv i t it) n I imi t , dent i Ii led Lo'., t l U J b-funde'd Even is i n the XD) cate&gory vari r d betLween a few hunrdreud dolil aiand $580,000. The mnediar Event co,,t wcI', found to be approxiiratel y$4 ,O000. Inidust rially furided Evenit s for t he same clas of re!,ea rchvdried ill Cost b l~weeii a ftew hun11dred dull a rs and $25~0,.000, witht a nied ianof about $33,000. So few RXD Eventb wer~e futnded by ot her SOU~r~cs thattheir analysis wuuld be statist icolly unreliable.

1.n tile two categories on which useful, -data uxist, the conkc~lusion canbe draw) thaL, although the. DOD can or will undertake far mrore expensivuresearch tasks than industry, rthe median RXD-Evelit costs are quite corm-.parable. ~Accepting the estimate that there is 'a dirffrence of less thana1 factor- of 2 between total DoD and total non-DoD investment in research,'th is f inding also suggests that the DoD funds a greater percentage oflow-cosf research than industry does.

4. Third Strategy: Determine significant management and otherenvironmental factors, as seen by theresearch scientist or engineer, that appearLo be commensurate with high utilization of.research results.

4.3.1I Finding: 'Tile utilization factor' appears inzsenvitive tocl&ia Jifferenceo in organizational a tructure or' profit motivation

L[I'lpettringj between U.S. .industryj, .in-house DoD laboratol'ies., and univer-si ty-aJvociated aevience and technology centers. It mdy., however, besnoitivJe to dliffer-ences between these types of organization and the

.d..VW~ic organizational ovtructure of universities.

Within these three major organizational types, the distribution ofRXD Events was found to be that shown In column I of Table V. The truemeasure of relative productivity of these three organizational d~ivisionswould require that the distrrbutlon of RXD Events *be compared with thefunds available for research in science a *nd technology in the sarrue timeperiod. While data of this type are not available, or are not known toexist in useful form, a less rigorous measure of funding correspondencemay be developed.-

First, over 91 percent of the RXD Events can be described withinthe National Science Foundation's "applied research" category; NSF esti-mates of DoD fund distribution, by type of performing agency, from 1963

to FY 1966 are shown in column 2 of Table V. A second measure of theI distribution of funds is available directly from the data gathered underProject HINDSIGHT. Percentages or the total cost of identified RXDEvents appear in column 3 of'Table V-again, classified by performingagency.

37

figure 12. IN-HOUS$E DOD- ACTIVITYPe es kl .PItve A. Pettentat' of All 1 vert's v%. I iwIm-

Doth~) lit11IP I-leh( hes b~t lit to d ter jqe of a~ 14. 310 __o_______

~it

20

51945 1950 1955 1960 1965Year

Number of Events S. Number of Events vs. rimeDoD In-House (Linear best filI to a~verage of 5.o 4..

p 100

800

60

400

20

1945 1950 1955 1960 1965Year

_X-3

I-~ t llt.It iii betIWeLiV1 RXL) LVIii all Itijdi~lj ti f iIi . ; ( Im l t i urn, I ui I but ht, ..v i I01)1 vc and L'Xlii'tled i tifiuti pert' f lse I O'. , W,~ Lýi I ciii citl illF CCII Cillilt', If Wiii

The 1 .11 io,% Val-Y onily % Itilit ly f roal li it. oryqinu/al loll tj~t-; In ,jiiutlii,1101 l t , LintdI at It' I th I.ft jitd i nq . I lie NSI~ i'tm it911 I 3lJ6')

10t Ile VX11LIC~t d to rep rewlnt the dlilri bUt i~l Of f ulth.l uve r t I0 v vi CPU-,I 8-year. Period; however, they provi de a tiseful fi r,.t -order aipproximiat ion.i

It iý t he na Lur i .eeac t hat dol lar aifotiouts culive r red (Ii Iy tScenlt i t i aild engi ive ring man-yea r% and his alI lows a second approac~h

Ak ~~~~to tilhfi tl mm rq .The tota I lat iona I populIationl of- sci en t istLs roug hlIyqulad I-Lp l ed durimg he pe r~i d f romt 91915 to 1963, whill e the same populationsegment in thle. DoD) laboratories is bel ieved not *even to have doubled. Ifthle Iifmnding is val id, ther fore, thle Dob laboratori es should 5how a gelt-eval decline by. a factor of approxim~ately 2 in the. pvvoctht.ap of RXDEvents generated, in-hou se and an I i crease by th e'same fa-tor for VIC:

:.tcnumber generated. Figure 12 displays the time trend of DO. in-house RXD0 Events, Showing very nearly t'.e exact results expected.

Available funding data for universities do no' permit an analysis ofthe relative productivity of two groups of research scientists-thoseassociated primarily with the educational portion of the universýity andthose associated with. the scientific and technological centers. Such

4 ~centers as the Lawrence Radiation Laboratory of the University of Cali-fornia, the Jet P~rop~ulsIon Laboratory of the California Institute ofTechnology, and the Instrumtentati.6n Laboratory of 'the MassachusettsInstitute of Technoloy ar urnly rceiving an amount estimnated at

less than 25 percent of all 1 funds for bas iC: and applied research goingto the entire *university, class-and, of, this., perhaps 60.percent is forapplied research. Comparatively recent in~creases ir. the-size and numberof these groups suggest that- their earlier funding was even less. Thus,it is significant to find that 75 percent of theý university-creditedRXD Events came from associated science and technology centers or aroseout of r-ecognizably mission-oriented prog rams .

Within the correlation accuracy of, the sets of figures. cited and[modified by the discussions, it is conclude'd that the utilization of

Lresearch results is essentially insensitive to differences in classicorganizational structure that may appear in u,'.s i 'ndustry,' in-houseDefense laboratories, and university-associated science and technologycenters. The finding cannot be extended to include the educationalportion of a university!.

4.

ht lit- vxiten't pe'Iinlitted Ily t1 , I t. .. I i iy uil the Itxi It PE'

form~er anid t he valiId it y of hi% rt'L J I I it,10rma I ion W,"-Y)(41 a h t eq1.11itlt

hli,, or itti nA mot ivatIiona. I e"', I hali 7J pe raotlnt ol the RXI) I yent E. I o... 1

f i ed a-. r W.11,L I, it *.: i unrze that - , I e',s thban o elid11 of I pt*r 1-1 'n AJ

t he total -appedtr to have cone atioutt bucatee tile pe rforoier wa', i itlr-c

esttld pri lar I Iy Ilit extend inq thle I nidsN or knOWl edge. tit 9$8 p)rC1ln

tile sc ient ifI ical By orient ed Evcntý. , I he scient i!,t appear-, to hovt loeeri

M'ot ivated chieflIy by his aiwarttnvsts tsa Itan actual probi-lemeiSiel

For thi5 part icul~ar ajnalyt-is, a dI~inicti~nW~ talled, iftut IItwI!CI

basic and applied research I n science, but between UlldIrVeCt.d (by ifn

agency external to tile performer) and 065sion-oriented -research. The

criteria for classification included the requirement that' thle pterfurlaia

ance 'be associatea with an identifiable generic problemn rather thatn (XIVt

Vague possible utility of tie result. Thus the data are considered,reasonably representative or the situation.

The distribution -within the missivon-oritented 98 percent shows that

73 percent was funded by the DoD, 19 percent by industry, and 8 percentby other Federal agencies or by foreign governments., Within the 73 per-cent funded by the Dob, 21 percent of the work was done as'part of an

organiLed program at a ufliversity-operate~d science and technology center,

12 percent as a separately funded project at a univer Isity, and 38 percen..

at in-house DoD laboratories; and tile remaiining 29 percdnt was done byindustry.

It is 'emphasized that the total sample size of scientific Events iso small that the subclass distriibUtions' shown above must be recognized

as being only roughly indicative o6f-the actual s ituation. 'The only con-

Clusion that can be-drawn with a reasonable degree of confidence, there-~

fore, is that most research in'sci ence Whose -results -were found to have

been used in weapon systems was undertaken as part of an organized-pro-

gram. Specifically, most utilized new sclent-fic information came from[4 organized research programs undertaken in response to recognized Defenseproblems.

44.3.3 F~indin As a x'ute, the most useful role'of ociance hasbeen tMat of prov tng phenomenol~ogical explanations to the cjtcr

*In all the systems studied, there were only two cases where recent

research in science suggested to the engineering commuunity a generically

t or fundamentally different way of reaching an objective. Those cases

involved the introduction of '(I) the transistor and other solid-state

devices and (2) the thermal battery. The other scientific contributions

were markedly similar in concept. The scientist, by gaining a reasonably

thorouqh understanding of a process, technique or phenomenon lused by an

40

AV

now

111&111i li t lfit' I It so he I ps d,:1 i V I t h. L 1 i[C 1111 rilc ' Al..g.ii l~ iiii

IU'C' I~ Ldaq'h 01w fud 1n t 1I IiI icapa ilty of dV'tec I i I r g or

fen .)tolv.t. impo li I iO)0 'tilt'. Jlorlnc. and niiit thi rgardj it ci ence i-

t er ldtechno logy. WIcr 1) th yin ithi defiit livon 69ýv Iperr (It of thIe resear-

1,1l11ia exapI nle ionsto the I engdA f iiD Le vI c Ir(.11L t i114

In~4 3tl 4 F ndin Poairaig .P/ en The fnamen2'~tal 1pivii linwty of ' d,-t -in or'

.aarhil cince. T~he taIicnt~iS obse rvtf ing! thema radar ,,I'j:h in .ctcx -

de i k ?l.3 Ileal.? kafldlkwkv ing.its p~ea rfj~orma 'nc ,adi hýre rds t.

B 1oecaus te here wasy a pau1L ityinL e Fini ofv nt ,s9peif tica Oy r thed teswardh

tol rcecncle ths a uie yPproent . IDi S)rcpncT therought a apoing asealyss Therim

e tiooiceinitanlobservtion a that thii e *niec salyrlisu

-u j scien isalmtific knllf a awllseutonsat Oh'slaw tl yqis

tBilitye crtherions Boyle's law of gvens dynamficsand o Newtton' wars.4ceie yet I t would reppeare that ayo the enierrmmesadues algebricalexprieste

Evonts, ghoeneraltsilutions from diffeentiel equaions, ortsimpleWA g mapdeto echnciqu e tha he spaenouterd in scoleg, bhruth doep fgaras no se .coneredwisuth th etails tfof t the unerltn theribe y. tefloigsaeet

Thue iseag therforme in whihe sciEentifimonowlaedge hihis aeavalablurerstanthdingineer, rathgeera thanilarthe c iharacteistics suheao te g oy

knowedinithat ibsimpotant was simiar oe nclnerusion l regaringtes ulosuseulc orho scientific knowledge israxels eqratnions. Ohmn o lrati, the y it

4 twul pea htthe engineer usesmasrtypcnd tsxt rlgrencas, theresulso-cniiAt or technilgiq lreesc that h av bnoneenclegen u codoesd andt pseemtconcerned

wit te hadledtabular of ghudraphical thorm.

Thsi stefrni hc ce.ii nweg smd vial

toteegnerte hnohr hrpeitc uha h g ftc

knwegta4- m'rtn. ASm lrcnlsonrgrigteIL-,

usflfr f cetfckolde sdanfo nosrainta

Io v ~i I~f I (r t W a . I I I I I it- deve I opitcr I tI1 11.1itid vi,~lly! vd r I iw j

eUlllfI. ' , d t ~he tate ofa kth n ibr t tffb ictat yn techiques i v olv11in itelec i

rdrceiIthe designsiAl~

powier ines. il, hanbrid brings, di getioer, iouprs fo ilttrsuf Iand ath-ennas.

In brt~ief ng inrsgid'ofr ai techniques and pragngo opoetaseul ct. utical

bhe uesedinswgeansstes ifrsarch tranfagerss wverqe aore concel rndabout-thes term inawhichs athenewinorm, atiniraeaalable tomrs deengrironees

thbif hi einr' ud ffr ai techn ology cand b patcal,

approahes t obj ectigne wasia~n reqsuireement ofd assembifiysyftemhoendi te adevince; ti-iedvcsfrmirraetasisol

Su 2) Thserain objgetiv what tore stsfyrcogie gnritcnologca rsequiremea nsytem or rsac aagr eeoecocre~

the~(3 frinwihTherewa inforidetifn.iablae objectivother thangntedersirtos forwardd the stat of the ar'

Even2-ts ialling intothe ofslirnt cateorylem rdeatfidiy adisncernible

thr greate dif gvir- tculty was ualycouonterasoied wit sepratin the s-eon

andthirdvategorfties peudgments oferesnecsaryh intcenilogwhat becast-rt

lgculd recogizemet aoreirmthwtdeieagnrc"eurmnt

(3Ihr a t dniibeojetIvoh hntedsr.442

0 it Al" A II k vi I , 61 pi -u ~I i d.l.~ I 1 i411 1~ 4- t .rit I rr j 4-dIIl.ii 1 I II i Iv ' oII '.* ki onl cItI I d Iv tIe i nl'tcili I, JI I I v IIIt: I hr t i iI ' di

I.1 i. V~ I S'/ ~I 3 q C i re e l1i ICI ''l1it) I Alvd by~ .4 )n' t.1 I~ It% w,,iuaoii y' t1 u. .11% d j let Ii itPI k I' 1.i vL I, If IND S (I GIt and (bu) It I pe.rt-crit t IlaI Wer II cl ahi i3 1 0 -,In t I~ *

I V1lth 01d 10o2 1 ii '.11 Ctll it fitc1 .y¾ Lt oI (3r opi I 3 cal ioun

C 01-1yin 2/ pvre -ct of I lie RXD E ven t ~.I he' i dent iI ied. objecct ive

01 Lh lre rsearch cflor t was; the sat i Is.Fac ion of- a yefle I-i C t vcc.Iui I o9 i cdlIreOILu I roe n tr . I ni a fow cas~es , pitI-t i cu arly Wbell I liv L Vel t. i 33W) vd it

I-id I Ir.2 or I- Idabr icaion process , the reslt !. ISof IIht! rv~a rc: we rc1`00111 tO hjve becIL- .11ppliedl) tio w t er i 'l- iir apI) i cat io oth0 ler t han 0101.1Uct illS i tie redL by thle res.ea rch pe rtortnrer Fur- exaflp Ic, the vacuumi-ýarc-'.me ItI i ng tecchn iques developed for Li taniuni and uranitiu' were oppliied a) -mos t d.i rectl Iy to steel prodUction, and the plasma-deposi t techniqules de-veloped for metal plating found use in the deposit jug of -refractorymnateriail. The number- of i dent.iIF ied exampl es exh ib it ing th is trans-ferchar'acteri st~ic. however, was not, big enough to wa rrl-int a. 5 tati St.i calIbrelakout~.

* In the remaining 12 perccnt of RXD Events,, the motivation was:

(I) a commercial end item,(2) basic rese'arch or(3) indete-int (uuly. for the reason describeabe)

.4.36 Finding: The pr'0a2'alfl of research in technobogyf orie,~ ted')6:0ur spe if (;t s of equipwment has been a pqzrtiaulcaŽ.? Succcelculu

a otp I:o a ch to g.snoia ting uti iized knowZedige.

This finding examines a noteworthy portion of finding 4#.3 5. Cer-tain projects or programs supported by the Miliary Departments wererepeatedly credited wi~tn, having been the' source of scientif ic or techno-logical knowledge used in other weapon systems.. Principal among thesevicre the Army'.s Missile AadMssile B prog'rams, the Navy's long-ringemine and. RETORC (research to .rpedo c;onfiguration) programs*, and the AirForce's NAVAHO missile program.

The common characteristics of, each of these ncl~udied the specificityof the program's operational goals and the freedom granted to performers

I-to explore multiple approaches to the solution of problems identifiedthiroughout the program. With the exc, epti~on of the NAV AHO orogram, nonteIWaSrintended to de liver a partic ular operati onal system at a given time.In every case management wbs centralized,' and apparently was able tomaintain a practical degree of balance among the ies cetfcds*ciplines and areas of technology in which further work was necessary.

143

A qolo al ki .1 L I ý1 in I n I ~ tit It At wn I fit t lit I itp1 "1 t-ts~ I tj ilJ. i .

iflolhV &It~vt I0npvd tilt. t ,:ntwuloi id I t -116 toi tII IL. it'll xila lit-, Iltt . iii I lit-d ni ~I tit v io tI it ild pro I I

# hretl apt1 id Lt iol of~ otmeIA tit It e.hno Inqy i limi lvd,P% piw if icity in dof ini ng Lit) ct:t ivci i c%'.Le'tet ialI W i I It i i t lit fLurl erktcapal) i I i iis of.~ p1lann i ng I. liods , thi. appell- r t; Inv he iBo-A reoid ii y Ln

tie -Il tilt:: r:2 a %ie 'c ripl ion of known " rvi Il wt.)r I Lit priibI -t

Z_ ~ ~~I ish a ialotgW~ ti l 'e '._ ivit i fic. andl A.chiiolog it~.s I 'CtsUlmull i Ie~i n vti wi lifIatgr~eletm.i is sought concerning 'tile minimiuit operat jona I capitbiIi ty t n.ltwoulId tio worth buying for at g iv n 'price and the ;iiaxiinuri 6pertti ond I Cap-abilitI y that 'tarv be -af forded at the pri ce witht foret~ait technology.

When agreement is -reached-and here the discubsionni I- limited toNystemis that are well beyond the contemporary state of the art-a del ib-erately advanced development program may be undertakt:n. Detailed design%of several potential-solutions to thle overall systeml problem can be pre-pared. The des'igns then provide the focus for a supporting researcti pro-gram. As long as the operational objectives agreed upon at the outs'et

S ~are not comprcwnised,:the scientist and technologis It can have the f reedomnof des ign flIexi bil Ity to -alIlow other -technolog ial compromi ses-. AlIter-

-native approaches to original design are essentia'l in order to minimizerisk and, at this state of the R&D cycle, are relatively inexpensive."Hetal bending,"1 or the conittitent to hardware, Js essential only to the

5ýextent that interface problems can be highlighte'd and identified.j-

-The value of focus ed prog ramn planning in science and technology isA most strongly supported by examination of the NAVAHO program. Contenmpo-

rary guidance and propulsion technologies, to' name just two, were Conti-Ipleteily inadequate to provide the first-phase NAVAHO missile, which hadia minimum range of'300 nautical miles. Althiough inertial, radio- Inertial',raldio- and celestal-inertial guidance 'techniques held promise, no ap-proach guaranteed success,ý Further'. the di fferent gui dance approachesplaced, d iss imilIar requ'i rements' on the mi ssi le Is flI ght profile-anywherefrom essential~ly constant-altitude alerodynami-c to fully ballistic--whIch -

in turn established requiremients for consi'derably different: rocket or jetengines, none of which were within the state of the art. In a similar-vein, dependenitly-variable requirements were established for other partsof the airframe and ground support equipment.

The Air Force approach to, this interpa o0rqieen wstprovide specific but very generous scientific and technologicall objec-tives, .and the relevancy of the science and -technology generated by the

A- NAVAHO program to other Defense problems is fully attested to by thewidespread use of, the new knowlerloe.

I. lht- .t i y k 1L Al Illy' M i'.,' i It'A .,fill M ,i i 1 13 1) 1 rjFý1 ' 1% qili tI.v

.i i iI of y Ilk' 0 l, 111' W ,tlf II:c IU t i hi a21.IC t Ie It lv Id by L IIIi ijC II -,y,

S I GI I Ali,)'ther OIV Ointati 1tl .1 1 I t, yp il.' L'(l I t)112I --().I itritedI ul1icelel

idteiiin, t. I at 1211 Ii n - tl1 4 , tb* iojl1iti n I )(L I y of 1 Vveil t le. I v d Ii IIJ to t lit. 11, 1m; id~ ')_ iliiir Ie . C 1 -I " e I y t ho1. t ct.' hiol oqy wq,, dc-lyt lopt-d Ii n u o (i~I(i I Ily

ash i III I, 0. 111 e iv I I Y.n 1 ly I 1-t Ile- I equ i I te kruw I edye )t li Ilij uVoi 1 11db Ivn3! 1I. ()1l,.i t wa, fiiei.dcd ,i[I t im' end iI I'mi.

Anot he r indi cj1tor of hie N111.1 tJ Ue oI oU~ed prog rami pI laiinig ri it -,L

ei11C' and ý echinology I"p royvided~ thirough analFyzi ng t he ra*te ol I' now I dgejic C iC 1.Off For t-his 01.11`001 thle studied sy.-ovinem' were divided ilnto I wi

()Those that went through a re-latively extenided preproto-type dod protot)ype& devel~opnient. path1-9r, at. least, for which a system-concept capproach Was adopted in the formulation of research and explo-.ratory-de-velopluent programs; -anid

.(2) Those that were dev~,loped in-an essentially once-stepA process, from technology base to final system configuration.

Typical exampl~es are the Mark 56 and 57 mines (Figure 4~) for thefi rst -arnd *the LANCE ,mi ss Ile system (Fig'ure 3) flor. the second.

All RXD Events associated with each system wi thin the two classeswere analyzed to determine when they had occurred, In terms of how longbefore or after t he system's engineering development began. The re-

A suiting data were displayed as in Figure 13. The potential advantageof using an end-item-oriented, systlems-analysis approach in formulatinglong-range programs of sci~e nce and technology *is clearly'sn'en.

The title of Figure 1.3, "Tec-hnical Confidence Level,"' was chosen topoint up another-advantage of focused planning. The RXD) Event is gen-erally associated with a scientific or technical problem (see finding

4..) hus Fge13suggests that there are typically three times as

many unsolved problems at the start,,of engineering development when thepreceding research programs in science and technology had not'been the.subject of focused planning. If it is assumed that the probability offinding timely, acceptable solutions to-these problems follows a Gaussiandistribution, the relative technical risk is nqt 3:1 but the Naperianantilogarithm of 3, that is, 20:1.

............ ..nretrospect, somre reasons the weapon-system approach to researchplanning should be profitable are obvioUs. :Coupling the research andapplications-engineering communities is thlenatural cons5equ .ence of theprocess rather than an afterthought. By being responsive to the Identi-fied needs of the applications engineer, the research community's Iin-

Aterests are focused in common w~ith those of the marketpl'ace for researchproducts.

45

. N.

Figure 13. TECHNICAL CONFIDENCE LEVEL

Percent ot new knowledqje utihiedo00

90

8 6%.

- .80

ADVANCED70--DEVELOPMENTS~70-

60 - I50a

"NO ADVANCEDi DEVELOPMENT

40 .i.

30

20

1-16 14 12 10 8 6 4 2 2 4 6

YearsV- (Time before or after date of decision on engineering development)

44-a

4

2$

-f

i 1 1I 1e', pect I ti Li. ll~ lii I I eiiiiiia o I lie i 1 .ll[).1 u1 tmý i..lir I ul ) I ii It 1fl ,

uId.0 iidI'L Ui ofI tq rj d * i l I It .i' ,,i (t..I o ul . '.()i I43 lit, 1. 1 I ll t*ip er tJi l ia (J~

li, ri% ffi o by Ilii. findinqo that ifielc> is a vy.hiqhl Corre'lat~ion.be tween O1 L i.'a L i oil o I -t-l'rcl re,.ul Is anid the fact that the iemr hadtf ri . I a Led thei p rob ICle . Cc r 1_1i r1y Ii !Uggc'st. that th11CU , (.c:I ItI J jt Ijij

i t y 1,01 dlc'f i n i icq a r'qu i Irellert L , i 11 molstI ca,.es the app Ii ca t oin%

those Events came from the group performing the research. In the re-maining 28 percent, technical initiative resided in the applications-eng.ineerinig group or was shared by both groups-or the informaption wasnot available. The dominance of the cases in which technical Initi~ativewas taken by the performing group is so overwhelminig and the data con-cerniingi the remaining cases are so indefinite that further resolutionappears to be pointless. The characteristics of the performers areconsidered in section 5.

During the pilot studies preceding those~of Project HINDSIGHT, itwas noted that successful pe:rfosrivers-in *the sense that they achievedutilized results-were equally successful in quickly obtaining necessary

fundjs and other cap~ital resources. Because this quick funding appearedto be atypical of DoD research support, it was suspected that the obser-vation was significant; consequently, a deliberate effort. was made duringthe Task I studies to gain quantitative information on funding delays.-The results of this effort were not satisfactory. In very few caseswjas there such a delay in funding thati the performers remembered andmentioned it-or this may be the consequence of a selective recallIdi fficlulIty.,

The Task 11 study personnel'are continuing to examine this matter.Subject to what would no~w constitute a surprising finding by the Task 11team, it should be concluded that, to achieve a high research payoffstimely accessibility of capital resources is equal in importance -to therecognition of a need and the existence of a source of ideas, togetherwith the. communication coupling between idea generators and users.

147

-1. ,,,.i fi) it"

Iiodillq 4.1.1 il' dim 14wl that I1 fit- tl-u Idtm tdt~ WW.~hi-I j wtii tItwo p i dl I voflP ilinIt P' Idli I tI liv w '.vart-I oii #,ii i v-,I t.1 the' Iiolitd itiiii l titokii) 111i'. illlpl iv' tlt- octitac lit11 Idn i A rl . e I lC

t~oi l.0111o1 i -dlituff, beivievti the Iva gioups~. Ava iili e dt .i do not de'tilt lt he eit I y 1e queit y Will) Witit il arly of they ',eve'ro i hI ic-

-4 .111oli I it i 0ii1h Ii hlki, ArV 'i lea I U1 ilul icdt itq requ i rvintid liw Bt helt datodo 'pre-'~ent a pa! I orni w!IIth regij di to 4Litwtoo it ii I iow. i it the oppo'. i ti' d i i cc-tiolii, tha lid: , 11wv p~rois of rt-st-t-rdi teul, ito~ ward tilt! evviltual Ir ur.

Table VI .o' thle f illdi ng' ill termt. of the scver il LldsSC'. of rt:-teircii or exp loratory dt'veiolopuwt pre'vioiusly devcr ibed, and rofv.itl',.Chivep yeneraI claNtst' of coummunicat iowi~ iritk': thle informal pt-rm-.#to-tpel-m~n Contact, tile published sttienti fiC or techinical re'port. -slid the

pro Oita I Sem ina r or sympot iurn.

Table VI. MODES or IDEA TRANSFER

LRX Iaeoy ___ cot:~ or report siltC"

Research (R) 45 53 2

Exploratory development:X0 64 33 3XD(design) 79 21 0

XD_(infg.) 77 23 0Note: Figures are perctie 'of te research class (horizontal rows).

Despite the fact that the professional meeting embodies scme of thecharacteristics of both the otiher links-generally a published 'technicalreport is presented, and there is at least an opportunity for person-to-person interaction-it is seen to bet the least often cited. Publicationis clearly the dominant mode for the transfer of scientific information,with a transition toward Informal., person-to-person commnunication as the

specificity of the technological Information increases.'9It-should be noted that the data presented in Table VI consideronly one attribute of the seminar or symposium. The true worth of thesemeetings is-not measured, principally because we cannot establish hlowmany valuable personal contacts resulted from encounters at such mee-tings.

J Giimore, Gould and others studied the flow channels of technicalinforiation utilized in commercial firms. The results they haive repo -rtedcorroborate several of the findings of Project I-INDSlGHT."

"JoThn . GilImore, Wil Ii am S. Goul d, e- ai-.,, Thu f'1d'.,i4x

it tog la ju tion *%n Colno p i at, Pirmu and tihIc : :`ls l,5 -zl cPr:'(Denver, Colorado: Denver Research Institute, NASA' CR-790. June

1Ali148

4"

Colvv t oi'. CollfLI It' Ir ' ,V % ymsup' a, a IId 1~s.1 '.10 111W., %1VI

11 .ii tllIy I'iliske A.11il " forSW I,'~( . i~ J' .U

'I'I t) I- I'~ alhe oppo r t Uri t y I ) uliet.I anldexchassqi' infioma t i on wi th t.l eaquv. ind t o inI'4)C (.v/iii

protICLc d i J. I ay'" . Manly illdi vi tliat CIUCqutti onedt insdi tdtLdthat. I armas papers probented at M~eeti igs tended ma Iinly to

siv lie, hitLeni ts of tht, !peake r (by b0oost ilj Ilk Iii, sa tU!anld t.hat K! 1,j t '?1 ii~'/d 'i.', I: i

I JtP~kII'l. [iaics. added.]'

1 he Deliver- Research I nst itute group further reports that:

Textbooks anid handbooks tend to be from two to five years'"obehind the state-of-the-art . Never!Lhe!ess, they

were one of the most important sources of information forproblem solving." [ItalIi cs added.]

The words, "or more," are emphasized hiere because, during the HINDSIGHTinformation-collecting phase, it 'was repeatedly observed that a substan-tial number of the most frequently used teXt references were I5 or moreyears old.

4.39Finding: he ei)ideklec dden , -cthe cotmenonly aocr-qed ik'yo' If-ýt that 1-Clatively fewJ orqanizationoi pirovide the majority. 6;, ti

se cientce and technoZogy.

Ar nlsi ftedistrbuio of* RXID Events and their rate of

production according to performing organization reveals two principalfeatures: Over half the identified participating organizations producedonly one RXD Event, and over,70 percent contributed less than three. Inthe low-rate mode, the industrial laboratory, was predominant, and as therae of Event production increase's the pri~ncipalognztoa otiu

tor shif~ts s~teadily-to the in-house Do0 laboratory.

Table VII illus'trates the process and shows that, when RXD Eventsper organization teach four or more, the:DoD) laboratory becomes a domi-

nant source. r

These findings are entirely commensurate With those cited earlierregarding the significant modes of information transfer, and they illus-trate the obvious conclusion that It is easier to communicate technolog-ical- requirements within a single organization, the DoD laboratory re-

Asponsiblefo~r' the system development, than'to disseminate the require-ments throughout a widely scatter'~d industrial community.

p. 25.

"FIb.id. , p. 26.

49__ _ __L

loi c V~~ VII RAII 111 ':, lk 1 lIVIN PRODWAMN. I ;I'I Y (P•f-ANI!/A I I OAI YHl'tI -- ! e--il (d i 01 ,'1 , ll,,It'!lhIt I ION .r i " ", L n;N M i J I 'V i ".

I ,I<. ¼, I.l. i-Ilrt ,y ( )I, iI,, ,,it ,-tt ) I '-j}

".FOne -(q A -II" 1kw' rtl or 1m re . , 1..

Four ,,o r, 3.,0 . I"Six or IW nore 5A)1 iglht or milre 4 .3b 19"Ien or more 60 201 20

SNets: JrreS art, Ltrct-i1t-il'; of "v'h rfi. e ;l,< (hior iz;onis. i (n,;,)* includes scienice and technolosly centers.

4£--Li.4l FourthIi St ra1eqy If Ihe f i dinjls uf LIi tiit' i.ra y idiLatc aSignt if i canit VC i iuni' " new 6L or t etLhlIUit (jY,devise a value-cost index (or set of indictes-)wi ici offers, 4 i tit ivtye mitre iof thie tel itt ri

: feethe ivenui>!, of thi

.'. {. ... m : mde---oss ible by Ltht purchasetd kn~owledge.

The f irdiigs relevant to this strategy clearly demonsltrate the re-iance of newweapon systems repon ehe results;of re!.carch in science, -r

technology to achieve an improved cost-effectlveness ratio. svctLn 7tot this report, "Requisite Level of Invest..ent in Research," gjives sotew*

ALI- qross estimates of the relationship between the cost of research and thetvalue received; but a more useful, precise, quantitative indicator fordescribing the effectiveness of investment in research.has not yet been

:a devised,.

"The studies made thus far offer some-insight into, the factors to beconsidered in an equation relating cost and value. The effort to devisethe cost-value index (or indices) should contintue.-

4L.4. 1 - Finding: A1tpltvpi uzes of scic' i ast tai~o~os :stt:zncot 2 iJ 1-! l _a11m1ttipliC'ity of weap.on 1y..,t~iies Pqi't*U rCI.c.Jtnr A/CC L1

Žeonhirvestment cost he appor*_t-ioned ai~mnon all Utili~z ins ai rethat a pcr'tiosn of the. totaZl Ialue add-d tr eaych ntil'izinj a tern :.vI. c.t-,.iba•e., t< the Identttfied Abant.

A given RXD Event may manifest itself one or more times inj a singleg. ir, -weapon systcm, and it may contribute to an increased cost-effectivenessi- ratio in one or more systems. Considering only the 20 systems that pro-vide the data basis of this report, at least 20 percent of the 710 EventsSwere identified through more than one system. During the -course of the.

investiqations it was frequently learned that the knowledge accruino froman Event had been used in systems other than the 20 under study. Infact, it is believed that, at the time of the study, over 80 percent ofthe RXD Events had contributed to three or more applications aside fromn

t -those in the study.

W50L~., s

tat!

'I. 'I.. I fill

Ilit I Vill I v I t it* I ItI I It- L 1

i ' (tj.I; i I iv rid v h i .p . d li t I iAw chI v. I I tui l- I IuIiht- I 1 1111 I it1, I ii*r s.nd itf i

11 ., 0hn lit'~ r 0 t i tII tn1 11) 1 tlt J!

ll 't' it;' 01 1 ti II dI Ot111 ',tlt 1 1 vi tI ~llt W0 i.hi,1h10 t1 Y'. I vi llpl.t ov 1 bt'.vf , f mn I ()III

,I o lit-% ~,- k u I ItI- - p'1L init M In lit' ht I Vi, ' L t iJ " I E. FI I OtI iV .- I 1t .1 1' It.- ()I ti t- ''J

Io W ilv. 1 011 [1yd itt li 1 ) .u t i ' A Ilt ntý I'l . s Id I qqI'., i. 1 p . 1. Il IvIdIiOol

r!starch)0101 Iil bei 1 14.4Ui'sul Futhr if all resear ich imosts arej to bv

oll fi.weaphn hysltem ltha t uexpt an pat of I I ihenw-A cieeo ehooy

This sl~ r intoue the naIotdion use tha valelmutio be copte intrm h

wecon-syste ed s mixt raterthniz a mutpe un or mdan gicstEfetventshos thationdth p roble apofc sesin e h const-valu te rltonacshipf" ain resevarchbecoe~searh wmattI er of peratnionsRfuI Furtler research. rahe tha acu ting. tb

4. Speifc ICd he all-eng ies stbish o thesignifiatonc ofs spcifintccon

Th einrodings toe no uesthtionsu ors chlenges raied by erisoftics

Management: of the DOD research. p~rograms.

The derivation of such challenges is described in the introductionto this repprt. .In this section, pertinent findings are discussed interms of the ch llenges., -which again are presented in the foni o yoheses.,

51

I F i i' t, Ow I I lengt* The OulýoV' requ irvimeni, I or jilii 111I itili H1 %C i ence 1d tI. t n W log

10y (.af be -I t "I '. ied 1o *,U(h .11vxten t by r'sea rch ..Upported by (fit Nd I i on.. Ar, :in or.I'

and Si~ace Adm in 1 .1roiti On Jid the o Nai 'm.I S et Flti oH hoI 01.1 %iii Canit reiluct on, can ble .made in theapplicatbl e port ion% (if tOw IDefie budget.

Less. than 2 percent of DoD-utilized rest-orch results., buth scien-ýtif-ic and technological, came from programs supported by other Federazlagencies, state governments and foreig 'gov ernet. Ti snt rprising 'in the l ight of the f inding that 61 percent of 'the RXI) Eventsoccurred as the result of a specific technological weaponi-system require-suent, and thalt 27 percent were the. result of research conducted to Sat-isfy recognized generic technological requirements of broad classes ofweapon systeiis. Df the remaining 12 percent of Events, 9 percent werethe result. of scientific research, most of which was oriented toward

'A i Defense problems; and 3 percent were for the solution of conviercial Orob-I ems.

Re latively little undirected research (as de fined in Table III) Isin fact sponsored by the DoD. The larger part was found to be veryrelevant to mission and therefore likely to be of primlary'value to usershaving qui te similIar requi rements.

Duin 94,the starting' point of the period-studied by 'ProjectHINDSIGHT, the Millitary Departments invested about $115 million in thecategories considered here as RXD. At that time the investment by other

)Federal.,departments was negligible. ýAs noted In Appendix F, militaryspending increased steadily and gradually to $1.5 billion in 1963, Comn-para~tively significant funding in these areas by other Federal depart-wments or.by Industry did not appear until late in the'1950s. Theirexpenditure rates increased more rapidly than the DoD~s, so that by 1963the other Federal agencies were spending about,$6.3 billion on scientificand technical investigations, and industry, about $2.1- billion. As aresult, for the entire period from 1945 to 1963, the.DoD spent altogetherabout $10 billion; non-Defense expenditures for those years tota-led $6billion.

If the process of technologi~cal growth were truly random-that is,independent of users' requirements-Project HINDSIGHT should have foundthat a much higher percentage of RXD Events were funded by non-Defenseorganizations. 'Because of differences In starting time and growth rateof the two funding categories, a ratio of 60 to 40 percent would not belikely, but it would be closer than the observed 94 to 6 percent.

This discussion and the ratios sugg2sted relate to all fundingsources other than DoD, rather than to the other Federal agencies alone.

CLg-t:jt' V ji j ft e'.. I,-# to ,itd woul h lIn t1: real i ;I t (it. In 'p ik 1Ilt. Ialtl.)Iy 'it I (I',t t ' IM )[nd ( .' 1k, tfdt tiio fit WC'¾ I f L' I I I Cu ~t- Mi'Id ~L PL f I IU A* are vr

L .ClI I I I 1 04 y h14t k illipor trifiL tot Oly iil~'.t' ti ited bvu r no ' i IIIi [I-Ii it y Ii I Ia tLot jilot.) t -ik ~e r- -but '.oc cit y tbrI CuI V Of I ciCIIL.t' 11, 1',. Idop t. I I I L 1', v won I If a'

J! Llene, C telii'11 tu i. ginJ i f'y t he maILny '.I it!clLL-,t Ofd 1,L dinoI0 og iCe'I, LIiu-'. La e-

Th us al1. thOugh. both Defnst!n, anrd NASA may be -,ai d t) Loniduct .iL'li r'C Laild tec0lin i11 cal inveSt i 9.Li uris in in a rca L erillt.d rucke~t p ropukIion,

I'Lr xafi~p I vit L ahnnot ,.be colic Iluded t hatI th c I-r work i %~ dup Ii Cat L v V(l0

tv.nuseul compl ement~ary'. One n11ight Suspect thlat, ýa- long as NASA i!sinferested ilrf cislunar vxplorafio'n,Athe related cquipmerlt and propulsionte'chnohnpgieswMilIl be qu it s IrIiarrto thlise aissoci ated wi th t he DOD IS

'Isa t eItIIIe and ICB M sys t eiis. Clea rly, oweve r. thereviil b1a1 eholgical divergence where 'NASA's intere~st, exiends beyond..the cislunar` intoregions in whih mtagnetohydrcodynaimic. and ior propulsion systemt. are

rat'it ca. Simir redcabeepected in communications and perhapsi n ~iaterilS', onlV Igen~eri~c names remaiining in coimmon.

SThe conicIu sion Ii s LhatL even greatly increased expenditures forsc ient If ic6 and techni~cal,_in~vestigat ions by Federal agencies other thanthe :DoD are unl ikely. to, resulIt in-satisfying many Defense needs.

VIP" It .52 Second ch leinge! The currently high-level support'of ýb~asi1c wdrk. is ,producing scientific and techrno-"Il6g ical informa~tion at such a rate that it cannot be

eff ectiviely digested, interpreted, disseminated or

:Put to practical use.

Finding: The available information is bcing adequatetil.. w i-roa.-zast but ja innuffic tent, despite the level of cxpendituries., to

oa kt ~jiqf all. requirements.

Were this allegation true to any significant extent, there would bea very low probability that the result of any given research effortmouild have mrany uses in different weapon systems. In fact, over.80 per-cent of the RXD Events are believed to have contributed to threeýýor moredifferent weapon systems, and those systems generally were.developed by

different engineering teams.

Figure 14 summarizes the time distribution of the utilized RXD

the dates of system design, engineering development, or first production

technological information is seen to have been generated after the nor-

moalizing date, and to a considerable degree was paid for with engineer-.ing-development money.

53

AU

,fl

4,L~

co N'

-.,.h JOJ.UM

Figure ~ ~ ~ ~ ~ ~ ~ ~ .,~. 14 ie i-rbto o tlzdRDEet

54'~ ~

I,-1 -till I -ic I, II I .I '1 ..oil I IL )It 11*1 J -w I I I tI 'I I iijI t!ilt ' , i I jkt imi iL it I ss td P I - ii I an,¾L lI i It Ii int- i ii~i I y li t i L)OM I Is i !1 41)1,II .1 -. i ~i I 'l i I-

1,11I1IL 11 Is lu l : 11,I 1 tin i I IICJ I Ili. t ( is Iii liY (A lIl tiuit. sIs * .s I I

1l il ,l I i -1,Y i 11 L 1, - uIs I 1 CL(i 1'.1 , ILilylC I 5s e

Ii ti I~IJ Iy I i e ll e IIIV oI Iillit~ t ill DoIe (-hiIlir I VwL41 11. Itl It t II

In tki' tli Vo,-. 1e1,1 () t~ lie effect 11c~ 4. ul th iii IdqLI I eer.~ d b

IhC oDO OVO r I. III mi I it LI Iy-Y",.c C11ncim hLudYt C01 L.'Sgry i b qucst irined , ýji /

Imnp roved 11dn11J1menvot and GritL ical stliect i en peroceduIrvm suchj-. those whichi dre being fl app Iio to IL CUIIn 1 x- and co'r ii y

weI pofl Sy',ttfims c(IuI d wel I ad profitably be extended to*the supporting efforts i ncluded. in the "'Mili tary Sciences'"*ca 1.0.9ry.

G ene ra I Iy L. the %ypergistic consequence of hund reds of i 'lenti fifablit.advances is the basis for a markedly superior weapon system, ratherthan ono, two or threo. important scientif ic or technological cont r ibu-t ions

In order to apply a cost-effectiveness approach, it would be nec-essary to desc~ribe fully each combination of advances that could permit

afuture't systeand then compaie the combinations. The.task would. be_kr further complicated by the multiple-us~epotential of the individual re-ý

search effort. The "shopping, lists" for all considered weapon systemswould have to be compared. before the most "cost-effective" researck pro-

gram ~ ~ -i col bderined. Itis suspected that the explense of suchaprocedure would exceed the anticipated savings, while the-time required

would cause irreparable damage to the program.I

A possible alter-native approach would be the development of analyt--ical procedures to aid in allocating resources amrong specific teclnnollo-

qis ht ar6-important to Defense. The-feasibility of such an approachgis sgetedbyth finding that certain technologies appear to be uniqueto the -DoD. It is reasonable to expect that the military services could

e: ~identify arid rate thoi r future reqlui red operational capabilIities . thatthose ýapabi I i ties could be reexpressed in terms of IFuture des ired n1il I i-tar'/ equipments , and that scientiIF ic arid technolog icalI advances to enabledeveloping-those equipments could be identified.. Then. it should betpo-sible to introduce economic marginal-return analyses to assist indecciionsocn resource allocation.

55

412VOt-X i II n Ity slit h a I) ot')cedIV JI))V r to be witIh i I th I ~ tv(. t I i - at I

P.of t he Ifltio olen ar t . I t% imp eilvilnl t ionl wolid bt e (1iteer.. l I y till'. iIenti

With) file Project IIINDSIGIII find~inq% I in rtjo~rd it) fill- %tvect ion tiof -

tind~obt ettly it) '11 act ivi ty ose Ia rgq a,. thte DoD rt-,varch pro~i .os,there art, somne Indly ma tasks. At lbestI, ltel sugges ted iaiagweltc t ap-proach Coul d clif mil a I e solltf t hem, at lite some I itut i i-t rodul. itl nq rit3k!,uLjqo-tce by Lthe fol tolwing ft n d iti .

4.5-3.2 Findinog: A 01". hw ,m',~ 11 ':-It z tq fi'.

The investigation of a sampling of the RXD Events disclose!, a*fairly commron pattern. Both the producers of the utilized scientificeffort and their cointettporary peers were aware of' the avenues of explo-ration that the others were taking. The disagreement by contemporary

authority as to the merit. of the approach that. eventually found use was

so marked that the successful performer broke with his peer group andsecured new funding. More than-anything' else,.these examples point out

the necessity for ensuring competition within bureaucracy an 'd the abso-lute essentiality of maincaining decentralized funding authority, or a

4$ multiplicity of funding sources, for every scientific discipline andarea of technology. The challenge-suggests a diametrically opposed move. -

As a result of decentralized planning throughout the time frameconsidered by Project HINDSIGHT, 67 percent of the requisite new know]-edge was available before it was needed. Since it is recognized t~hatthe saving to the nation as a-consequence of this new knowledge exceedsby orders of magnitude the total cost of the research program (see sec-tion 7), the real measure of the ef fectiveness of research management is

the degree to which new knowl-edge is avt3ilable when nee~ded.

4.5.4 Fourth challenge: Significant portions of the

near a point of marginal returns.I

The analysis presented in this report is based in 710 RXO Eventsidentified by studying 20 weapon systems. It is estimated that a moreexhaustive study of only those systems would have revealed an additional400 to 1000 Events. The median cost of an RXD Event was $45,000. Anaverage additional cost factor of 15 was found necessary to carry the

median Event from initial demonstration of feasibility to readiness forapplication.

Recognizing the weakness of these! figures for anything more thanorder-of-magnitude estimates, and assuming that there were only 1000

rEvents inI all, One Ca.A place the cost of the RXD Events for thes.e 20systenis at approximately $600 million out of the estimated total ofSID billion invested by the DoD In research between 1945 and 1963. The

56

I Ifa. I it II I .Y.iI't't IIdVk* tiLLv.~, uii ty okl t I lt-e I-.t ofi LQ.ý

th lIit- I 5L)~ Ca, V 01 C IL C i.0tf1 P I 6ULedLII(' ý,LJyg it-t ud 11 1 lie 'di ,,L u .i ul

J, fj lid i Ii, 4. 5. 3. w~re 1 t U 'bC' dtvVe I tipel k dr( i %pIei' e ubrwa rg riau I.reC -W1Ic: I t Ci ' 0o LId b ILe i Lit-lt iI'i ei Ii :1 a tll~l t CI s~ % UbJ t. ye- fi l 11on

4. 5., F i, r I h'c lit, I I ciige: The support (A s i ent if ic researchfu dn th i~vers. bigtnIfei'ba1lit.o Iandtt thr~e, viabltfnhi und aq versIe~ by ni a th Iojy an 17ftct curre. viI t iI

.01 t he best: fI'a. 0 t y LaI el it to each S tude ntLs

Obviously., o'ny activity undlei-akeni'by a professor other thanteiching or counsel ing the -student body coul-d detract f rom h is primarywork -.imply by reducing the time available for it. But i t is gene:rallyagreed thajt the professor must continue to do-or to d irect- research inhis chosen-arL'a or ri sk haVing his scientific-expertise rapidly diminish..Thus the real concern'(a~nd this may'be+ Inferred from the challenge) i5

* whether or not DoO-ýsupportcd research activities at a university have,in the balance., a deleterious erfect ,on the students.

ýThe investigations under Project HINDSIGHT do not provide a directand unequivocal answer to this challenge*. But many-of the findings,

* reinforced b-' conclusions of other relevant studies, offer usefulinsight. Findings of this Project that bear on *thi.s matter (devel-opedmore fully in section 5.3) are summarized here:

Most important, the HINDSIGHT data show that the universities re-ceiving the greatest amount of DoD funds for research graduate a dispro-portionatly high percentage of the people who later produce the scienceand technology used by the Departnientof Defense. Specifically, the 23universities responsible for doing most of the research that the DoD

* supports in academic institutions aw4ard only 25 percent of the doctorates!6 earned in the United States, but that group includes. 50 percent of the

Ph.D. 's who have been Identified as important contributors to RXD Events.Similarly, the 23 universities most heavily supported in research by the

* DoD award about 11 percent o~f the Master's degrees in science and thearts in the na tioniý but 46S percent of the identified contributors re-ceived their degrees at these schools.,

The findings of Marquis, Alien and de Solia Price,6 based on theirstudies of the postgraduate education of the country's engineers, are

f.GMaquis and T. J. Allen, American Ps.hloit 21 021 961' 1err4 d. df- Solla Price, "is Technology Historically Independentc). 3,.ience: / Study in Statistical Historiography," Tecimo'logy cvng

* 1t1-* --: , It. Fi,1hi 1965, pp. 553-568.

.5 57

"4

-. thoroughly consistent with the foregoing observatiolns, anid clearly !,ug-gest that ýthe engineer continues through life armed primarily with thescience he learned as aln undergraduate student. Tile greater his profes--sorN' knowledge of the state of the art, the more advanced was hisscientific education. So It is not surprising to find a possitive car-relation between the level of the DoD's support of a university's re-search and the subsequent useful productivity of its graduates.

The time elapsing between an engineers' graduation and the peak ofhis productive work-on the average, about 10 to.15 years--seems to beindependent of his scholastic experience. All the patterns revealed in

6 this study suggest that it is a function of the rate of the individual',scontact with actual working problems. At this distance from his lastsignificant exposure to science, the engineer has learned enough abouttypical problems and their technical solution that he can solve a newtechnical problem on the basis of that experience and hiv-knowleldge ofscience.

In brief: All observable quantitative neasures deny that there isany great substance to the challenge that DoD7-supported research inuniversities has a generally harmful effect on the students at thoseinstitutions.

'i5

!> 58

[:1

a 5

ADD 114ONAlI IiNDlI N(G

IH pimr da11,1- I)t J host, Of -Projvct [II NOS s lcm-cors i.ýt of 71 0 d is( ride tEv01nt de-se ii pt i 011-n thtpr' .1V'L.rW1 befr~ilat (!,Vt Appendi x C) arid Yi.Jdk-

le sc rif I i't i n,; . 840re comrp I etIC in al e s sent ial di.ta 1-; , i' 26 are. adtquatvto 0support oinly a port ion of the aeu 1 y'ses. The f.1 nd i ngjs presented in

Lct jion 4 are t ho,, v cons idel red mios t re Ieva it to the quecst ionS (out I i ned Inr

sectilons I adýd 2 of- thins report , bot- the aiial1y'ses on which they we rebased d 1(1 not, exhaust tho potent i a s of the- accumulated data.

In*'th~is, Sec~tion and the nlext two, this,. jif-ornat ion i b. coib ideredf urthe r. Firist tI he data are diVspl ayed, and whe re pos s ible correlIated

with the f indi ngs of othe r reported researches , and sorte inferences are

drawn. 'Section's '6 and 71 contain add itIi ofal i nferences regarding the

n1ature of a high-payoff reseairch environnIrentand future Defense require-

'men t, for research, funds.

5.1. Qual i ty of the Data'

5.1.I Uniformiity: Despite efforts to maintain uniformity in the

data base, there' i-s a conisiderable a'mounht of unevenness in the more sub-

* jective areas. - P~roblem~s Arose', for example, i n separa t ing-ýthe reallIy

inventive activities that advanced *sc'ience and technology from more*

-routine engineering work barely withiti the state of the art. Consensus

Of the study teams was the primary. criterion for screenting the activities

in this respiect. A review' by the Project Director fol'lo'-Adl,'resulting

in the rejection ,of 2 'to 3 'Percent of the Events surviving the first

scrutiny. Undou~btedly, another independent review would have somewhat

different: results.

Also contri~buti ng to the data's lack of uniformity was the diffi1-

culty met in ascertaining the precise terminal point of an Event. By

definit~ion, an Evenitterminated at the instant feasibility Was first

demonstrated or valiidit proved in'practice, Judgments about what con-

stituted proof of.validity, differed wide'ly. Because this problem wasnever saitisfactorl ly resolved, the distributions of Event duration and

cost were viewed in the data anal ysis as only generally 'indicative of

the situation.

5.1.2 Nature of the Events: Although each. E vent W as considered

unique, a preliminary anal ysis suggested that it would be poss;i blIe to

Cla5Ssfy Most. or.. the Events by type with a relatively limited set o

descr'ptors. To this end, 15 descriptors were selected; anvd for' each

Event it was statled, "This Event advanced th e 'state of the art 'by 'the-

application of to for ,"employing the three-ap-

propriate oescri.p-tors f rom ýte following l ist.

59

"DIISCRIPlORS FOR CLASS11YIN; IVINI'

It~chNew I mu., I ior li M,) t heina.t i C.. I ',tabi s Ii l~P ~ f c io(ll

HNa tura I plnut'nd Mnufa. In ri I.q (pI cL-",%Spvci fic funct ioni LquipmentNew technique AnalysisEstiablTIshed technique DC!,ign 1 peCif iLdt ioil

New nmateriil Theoreticl vilidatibnEstabI ished material

The consequent distributions were dnalyzed, with the followingindications:

5.1.2. I Scientific Events: Ab 9ut eit 9n jv ' 1i of:Ott

It should be noted here that a very rigorous definition of "science"is used in this report (see Table 11). :For example, it is not synonymous"with the DoD budget category "research," although there is a usefuldegree of similarity. tn interpreting the significance of the 9-percentfi'gure, however, it is more important to know that, even though the

A Event occurred in a university and the performers might normally be_A called scientists, if the nature of the work tended more toward technol-

.0 ogy than science the Event was categorized as technological. This par-ticular finding, therefore, says nothing about the relative value ofscientists vs. engineers or of an organization's science and eng, -tringdepartments.

Recently the Department of Defense budget for research has been

roughly one-third the size of the one-for exploratory develtopment. Ifthere is any relationship between a program's size and the number ofutilized results, the 9- to 91-percent ratio of scientific to technolog-

.1 tical Events-rather than the logical 1:3 ratlo-suggests that, in prac-

tice, a considerable portion of the work funded by the DoD as researchis in fact ,technology, not science.

5.1.2.2 Materials: The development or application of newmaterials was the subject of slightly over go percent of the E6vents. Ofthosee, about 3 percent actually resulted in the creation of a new mate-

40 rial. The remaining 97 percent dealt with the exploitation of a newtmaterial Is characteristics (a) to permit the use of a new manufacturingprocess (fl percent), (b) to enable the development of a ncw device(16 percent), or (c) to advantageously replace an older material forreasons other than manufacturing economy (68 percent).

For, example, the last category might include the replacement of

germanium by silicon in solid-state devices, or the use of aluminum in-

stead of steel in the carriage of a howitzer.

In general, this finding clearly demonstrates the imiportance of

materials research in the advancing of technological capabilitf. The

60

!-g

J

4Ldis t r ibut ionls wi t hi 11 this 30 perct-lt of Evelit, howevur , suqgi:-t that(eff ur ts devotedI t~o yain imn .3 Completie unde rs and i Uyo icar~ir

C is , o ic5 of new illa tCrIl 11nd t0 enVSU ring Lthalt l knowledgv was, widelydis tr ibuted throughout t he t'ng ilneelr iiig comuniluityWUd eoru'u1

ha.nl attempts merely. to scvk another, somehow i ulp roved mla te r ial bv ious- ly, a commiensjurate r ecommenda t ionl ou I d be %tated i n termrs of

I .relati1ve emphasi rather than a choice ot alternatives.

5. .1.2.3 FunctLions: In determining whether the consequenceAof an Event Was the ability to accompli1s~h a new f unct ion , as opposed to

improved ways of performing an established function, it, was necessaryto use def initLional cr ite ria.

In general, if an older device or tec Ihnique. could have been usedto achieve the operational characteristics set by the engineer-.even ata h~igher cost or With slightly less reliabili ty- it was hold that theEvent merely offered better ways of executing an established function.Conversely, if the Event resulted in a higher level of capability thanexisted before (e.g., greater power output by a radar transmitting tube,higher specific impulse of a new propellant), the result was considered

A to be a new capability.,.

Wit~hin those broad criteria, the invention of devicea, or tech-niques, other than through the exploitation of a new material, accountedfor .33 percent of the Events. About 75 percent, pf these made possible anew fntion or offered a new capability;,-25 poitent offered improvedways of accomplishing an established fnction.. These figures suggestthat, early In the engineering development of a system, about three timesas much attention is paid to meeting performance specif ications as isgiven to lowering cost or increasing.the system's reliab 'ility. The datawould therefore supor the hypothesis that, at tetm f nierndevelopment, most~advanced weapon systems-are.at or near the technologi-cal. state of the art.

5.1.2.4~ Design: Inventions, or conceptsj involving ap-prahes to system or. aubsystem design were the bases for a little more

*than 4 per'cent of the* Events, or an* average of about 1.5 EVents, perjsystem studied.. .

ZJ1W In the opinion of the study teams, this suggests that the funda-mental concepts underlying these systems and major. subsystems did notdiffer greatly. from those.on which other existi~ng military equipmentsof lower performance were based. The main conclusion--is that markedly

w4ý advanced new weapon systems emerge as a consequence of the skillfulselection. anddintegrationp of many innovations from diverse technologies,wh ich comb ine to produce the performance demandrd. . It is junlikely-thatsuch weapons would~come. into being as the result of the mythical "greatnew idea."

61

'A

2.•PrIt, Analysis: Ahuott Ill percent of tlhtV Lv( ntt A evidence a differentt not nmessrily untexpectutd lhwrc•t eriili(. j

they appear to cons i•,t of aunalyzing a problelm in) de-tail .1d coupP iieq th liere¢ult with an alllostl nkwrpholop icalurvey oh ov• i(l lab echnoug c., inf

order te find a useful UsoIlution. Nonte of these [ventL •la bl dscLiibetlwa. ., peri.d of creativiLy followed by the test tf an i de.i that caiime iniiobeing duoing that period. Th e ingenuity observed in this cia•s-ofEvents 'is revealed more particularly in the framing and analysis of thi-

A. problem than in I lie' evenfual solution.

Philosophicallyr, the activity described would seem to --Illow thecammon concept of engineering more closely than the findienb 5.1.2.3)that some 33 percent were inspirational inventions. I If the data hadbeen structured somiewhat differently in this portion of the report, manyEvents included in the discussions of scientific research (5.1.2.1) and

materials(5ý.i.2.2) would also have consisted of analysis followed bysolution.; In fact, that-may be a more typical description of an Eventthan the one chosen for the HINDSiGHT study (see Appendix B). It cer-tainly is easier to understand the high cost of scientific and technicalresearch when one;appreciates the tremendous amount of analysis and

search involved.-

5.1.2.6 Other Events: Finally, II percent of the Events

lack any obviously common characteristic, except that each involves

some amount of creativity or ingenuity.

5.1.3 Content of-the Events: The Events can also be'usefully cate-gorized in terms of the primary scientific or technical area involved.Since the weapon systems and equipments selected for this study are •gen-erally typical of those that are Important'-to the'DoD, it may be conjec-tured that a ranking of the scIences and technologies, in terms of thefrequency with.which Events occurred in each, would give a measure oftheir relative importance in,' at least, the immediate past. The Fieldand Group Structureiclasslfication code and descriptors of the Committeeon Scientific and Technical Information (COSATI) were used, with theresultant distribution shown in Table Vil.

It cannot be concluded from these data that electronics Is moreimportant .to a missile'system than, for example, propulsion, navigationor communications. In the absence of any one of them, the final systemwould very likely be inadequate. Because each Event involved skilledpeople,-however, the distribution In Table Villi suggests the relativenumber of scientists and technologists that would be required In each ofthe areas listed to maintain a balanced base of professionals.

"9COSAI Subject Categry 'List M(DoD-Etended) (Washington, D.C.:'Defense Documentation Center, Defense Supply Agency, AD-624 000. Decem-ber 1965). [COSATI--Committee on Scientific and Technical Information,Federal Council for Science and Technology.]

62

-4

Tabl L VI I I SCIENTIFIC AND TECHNICAL AREAS OF [VENTS:FRE(QUENCYýQ, (OCCIJRRENCE

Area.

Electronics ------------- -------------------- 25.2Ordnance --------------------- --------------- 12.3Missile technol~ogy-----,----------------10.5Propulsion and fuels------------------- ------ 9.9Aeronautical science------------------------- 9.9Materials science--------------------- ------ .7.7*Navigation and commnunications------ ------------ 6.4Phys ics.------------------- -------- 53Mechanical. and civil enginieering---a----------*ý3.2Methods and equipment---------- w------ -------- 2.1Behavioral science--------------------------- 1.9Chemical technology -------------------------- 1.7Mathematics-w -- - -- - ----- - - . 15Nuclear sci~ence----- ---------- 0.6Biological science-___ ----- --- -- 0.6Atmospheric science------".-- ------ 0.4*Militar~y science ------- ----- 0.4.Energy covrin------------------0 4.

5.2 Motivation for the Events

In this study, motivation is defined in a very narrow sense. Theobjective here was to ascertain'what' recognizable influence, if any, leddirectly to the creat ive. act' identifIfied as the start o~f an Event. Ex-cluded were such matters as the reason the pefre a.iapsti,to do the'relevaint work, 'Whether the effort was sought of his own voli-tion, what perso cnnal drives may have.ntri~b teiJ to his motivation, andthe like. Iii this way, the task was c~onsi derably simplified.

5.2.1 Kinds of.Motivation: The "Permitted" categories of motiva-tion were those shown in Table IX...Definitions, of directed and undi-

*rected research are gi .ven ihnTables 11.1 a nd :I Ve. Alternatively, thesecategories'might be more adequately describ~ed as phenomena7-oriqnt~ed andapplications-oriented, research-but only if It is clearly understoodthat neither orientation'WoulId exclude Investigations of a basic nature.Within these definitions, the eight very important Events in the history

* of the transistor are classified as directed research because they alloccurred at the Bell Telephone Laboratories,'which is obviously appli-cations oriented.

63-

%

Table IX. DISTRIBUTION OF MUIIVAIIONAL LAT[LGORILS

qategory of otivatiorl

Science:-" DYie ted research (DoD) 7.0

Directed research (non-Dol)) 2.0Undirected research (0.2)

t Technology:

Generic (DoD-oriented) 27.0 +2SAdvanced development or system concept, 41.0 •2_

System in engineering development 20.0Not DoD-oriented 3.0

In generic technology, the.attention 'of the performer is directedto a broad spectrum of functionally similar, general-purpose technologi-

.4 cal building blocks. Examples of this class are general-purpose transis-tors, ferrous and some nonferrous structural materials, liquid .and solidrocket propellants, exploslves, and printed-circuit boards. As thefocus of the performer's attention is sharpened by the consideration of"possible specific applications suggested by a particular system conceptor by a system actually In preprototype development, the motivation forhis efforts falls into a new category-advanced development or systemconcept.

There is an overlapping of generic and system-concept technology Inmany aspects, especially when, the same people are responsible for bothsystem-design and the sponsoring of growth In the relevant technologies.For example, there are regions of technological overlap in missile guid--ance and radar in which it is difficultto assign an effort Xo eithercategory with any degree of certainty. Further, theperformer oftenfinds it hard to unequivocably trace his motivation. Thus, from 2 to 3percent of the Events shown as generic and system-concept technologymight be shifted.

5.2.2 Other Motivational Links:' In addition to determining'theperformer's motivation, information was sought regarding other, non-psychological factors that might have directed his attention toward the,activity that culminated In an Event. In a very crude sense, the moti-vation was the objective that guided the performer, whereas the motiva-tional link is the reason he was so guided.

Table X presents the distribution of Events among the eight mostobvious links. The three percentage columns show the sequential particl-pation of each link in an Event involving additional acts of creativity.Of the several categories, cuntractua) requirement, routine area assign-ment, and spontaneous Internal conception are self-explanatory. In the

64

"tol i lt. th' p Illi)rIll.,r5 wwr inlorii .d of 0he tuX;I. vLncU of d vd, ul0 'j

poý'Niblt -iolt) tionl t'or •fl a0"-'dy idelnti ei d problem, by way of the in-di c tt0d trjIIsIfI Ii Ik.

* Table X. MOTIVATIONAL LINKS

Category Primary (%) Secondary (%) Tertiary (%

Contractual requirement ------ 11.1 30.5 38.1Routine area assignment ------ 47.8 42.7 33.3Technical report--.--------- 14.8 18.1 16.6Professional meeting- ------- 1.5 2.2 4.8Extraorganizational* person-to-person --------- 8.0 5.1 * 2.4Spontaneous internal

conception ---------------- 9.5 0 0Analysis of foreign

product ------------ 7 ------- 0.5 0 2.4Newly hired personnel -------- 1.6 1.4 2.4No link apparent- ----------- 5.2 0 0

The dominance of the routine area assignment among these categoriesshould have been anticipated in the light of other findings of thisstudy. As noted in connection with finding 4.3.7, in 85 percent of theEvents the creative activity was in. response to a problem brought fromoutside the performing group. Where more logically should a problem be

- taken than to agroup of professionals who are known to be regularlyworking in a tec~hnical area in which a solution is likely to be found?And what most probably would be the nature of a*solution offered by agroup that Is active in certain technological areas?

In describing the production rate of RXD Events, Table VII (page

48) demonstrates that a mission-oriented establishment (the In-houselaboratory), if at all involved in an Event, tends to be deeply Involved.Apparently, having developed expertise in areas pertinent to. Its mission,the laboratory Is brought relevant problems by outside organizations

.,that recognize this expertise. .-

These findings form a consistent pattern and consequently offersome guidance to management. They suggest that usable ideas, are morelikely to be produced if an organization deliberately establishes areputation for skill in a limited number of problem-associated technol-ogies, instead of diversifying its talents over a wide range of techni-cal fields.

.• 65.,'

r...i

3 Fundint of Evcts :

t it F. 1:.•Ia ti onf aN githerecd with rei•arq d Ito Itwo v.pectý, tof Lvent 1 nd-ing : (1) Who furnished the money? (2) Approximnat lely hIow much nunltey wai."I nvol Ived? Ani Event was res.tricted by definil ion I*o tlt "initial delliun-

stirat ion of validity or. feasibility" in the hope of keepinn the numberof fund ing sources to a manageable minimum, but even so foul Iiple fundiri9was found to be a coitnon occurrence. In fact., 20 percent of the Eventshad at least two simultaneous or consecutive sponsors, and 6 percent hadat least three. Dual funding by consecutive sources appears to haveresulted from one or more of several situations:

S(1) The idea occurs in an organization that is unwilling to

fund more than the most superficial Lest;

(2) The idea occurs in an organization that, lacks the techni-cal ability to accomplish the requisite tests; or

but (3) The idea occurs in, and is tested by, one organization,'uta second organization which Is urwilling to accept those test

'PA-' 'results funds a more detailed project.

The simultaneous funding by-multiple sources invariably happened-A .- when two or more organizations were interested in pursuing an idea and

agreed to share expenses or talent, or both.

Table XI shows the frequency with which the several identifiedfunding sources appeared.

Table XI. FUNDING DISTRIBUTION FREQUENCY

Source Primary Secondary Tertiary

Department of Defense --------- 574 131 37Defense-oriented industry:

Specified-industrial R&D--.. 18 1 0Specified profits from (7

other DoD contract ------ 3 0 0"Unspecified ----------------- 51 2 0

Non-defense industry ---------- 35 4 2Non-DoD Federal-------------- 8 0 0Foreign and miscellaneous ----- 9 0 0

Table XII reduces these data to participation percentages on the"- basis of number of Events rather than funding level.

r, 66

I

1.111d i 'It SOLI -C V

Ih'.)drihner't Of DOe eir. .B4. Bileferst- r1 onr t ed n dtjs try 8 .6Other indUSiry 4.b0 th er F ode.ra I 1.0Fore i kII arnd ml sco II d nous' . 1 .0

Fa ch of the 28 Eve~nth in which thcre wa% simultaneous multiplefunding involved the DoD as one contributor and defense-oriented industryOs the..other. Distributi~on of consecultive funding participation bysponsor 11ay be inferired .froqm Table XI. 'Where the initial support camnefromn eit;ie'r the.DoD ordefen'se-oriented- industry, the' follow-vjp supportwprovided by the DoD., Where non dfnse industry furnished the

initila sup port, it tended also to give the follow-ups support.

fromrno? (te;~ rdta h efomrsslr for a few days up to a

mate, in terms of men, materials and overhead, a dollar equivalent that

Because these estimates 'concerned the approximate cost of incidents.n that had occurred as much as 20 years before, and because there was

little uniformity in the way the many study participants applied ther ~definition of an Event, an average Event cost is not a particularly*

meaningful figure. Instead, the median cost of anEvent is offered asbeing more descriptive of the situation. Table X1I1 gives median cost.Figures for the dominant Event categories, and Table XIV displays Eventcost distribution, regardless of source or other factors.

Table' XIIL. MEDIAN COST OF EVENTS.

Category Median cost

Sc:ience $60,000Technology (DoD-funded Event) 45,000Technology (industry-funded Event) 33,000

67

Table XIV. DISTRIBUTION OF LVI.NT COST

Cost range--

Under $10,000 17.7$10,000 - $50,000 32.6$50,000 - $100,000 19.0$100,000 - $250,000 17.0$250,000 - $500,000 6.3$500,000 $1,000,000 3.5Over $1,000,000 3.9

5.4 Idea Protection

The flow of ideas, as well as their application, may be restrainedby controlling the information In three ways: Assigning-it a Defense

-•security classification,.labeling it proprietary, usually to industry,or subjecting it to the patent process. The incidence of these re-straints among the 710 Events is shown in Table XV. Since it is possiblefor an idea to be patented and at-the same time classified as Defensesecurity information, the percentile figures are not additive. Nor is"Special Handling" additlve, for it is a subclass of both "Secret" and"Confidential.." The number of Events that are (or were) considered tobe proprietary information by the performing organizations Is not known.

Table XV. IDEA PROTECTION

i -;Restraint %Defense security classification:

Special Handling; 2.5

Secret 4.6Confidential 12.2

-Patented 18.0

Recognizing that th;s study has identified most of the importantscientific and technical advances in weaponry and other military equip-

ments over the past two decades, and in view of the current European

concern over the so-called technology gap, 1 0°these data are particular-ly enlightening. To the extent that such a gap exists, Federal security

s 10H.R. Lieberman, "Technology Gap Upsets Europe," New Yo6k Ti•4•+,12 March 1967.

68

ColltrolI ove r the H1ow of i tifoUrmint ionl i ~,obv iousi I y 11 Lo a s i gr i if i Cant L0on-t r i but o..I n .83 pe rcenli . of tile cases there, wJs folu egal Lcmnst rainft on)

open pUb!I i cal Li on of the i n f o rma t i on.,

The data, of coursie, do not ifeasure thle effect on inforiiation f lowof const~rain[ts imposed by Defense Contractors, either- by discouragingtheir people who inigiht desire to publish or placing del ib~erate. restric-tions on proprietary: information,.

5.5 Partic~ipating .r~ganizations

Inthe-same way* that an attempt was made to curtailI f undIring sources,the definition of an Event was deliberately fravined to. minimize the pos-sibil,'y that a number, of organiziltilons would be f~ound to have partici-pated in *any one Event. Again, success was noit complete. Only in 78percent of the Events was all the technical work done within a sit)gl6organization. In 17 percent; of the cases, two organizations were iden-tified, and in 5 percent there were three. Where multiple performingorganizations were identified, the combinations included the f ollIow ing':

Government Ilaboratory / Government laboratoryGovernment laboratory /universityGovernment laboratory /industryinidustry /university.industry/ indus try

Thus it is not' highly illuminating to discuss the- ds t ribut ion of per-forming organizations in terms of percent of Events by organizationalclass. Instead, Table XVI presents the incidence of participation inEvents as a function of organization class..

.Table XVI. INCIDENCE OF TECHNICAL PARTICIPATION.BY CLASS OF PERFORMING ORGANIZATION

Inc idenceOrganization class.() -

Governmentl laboratory_------__'------- ----43.6(DoD------------- ------------------- (42.3).non-DoD ------- I------------ I---------- (0.9).-industry-operated---------------- ( 0.4)

Industri al- ------.---- 44.2profit ------ (42.3)-not-for-profit. ------------- 1.9).

University-----------------------.1.0i 6academic institution ------------- (_5.4)research center --------------------- ( 5.2)

Foreign laboratory (all classes) --------.-1.6100.0

69

Since' lib]ie XVI cover!" thti elt il pel lud Ifult.Y~iu ~ ii.not .1 meatsure' of t he- cut ren t e I-1t i Vt ilpli~ ilA (i~t litt-!, fit /dltlt irv,1;j

c las.e . 1 he Irenid I it rela tivye imipo taict - of. tilt Cver illitt1 t -lA itJand i adiS try, 0h0 Lwo pr i mary cnt r ibutt r- ovvr t he whoI e I ifilt: I raml:. 11ilust ra ted i n Fi gure 12 {Page 31).

Sect-iv -i0 !3. 1l oted (flat about 15 Pecwflt of t-fhe llli veril Y-creite RXI Lvrjt, cmefromn ztii.ociatedl scienice and technolugy LL[IteI

or arose out 1'I recognizably missioit'-rimted programs (g.,thie irtves,Ligat~ionS it -missilIe base drag charactv. ikt ics, Event No.,0060, undertire direction of Professor Korst, University of Illinois). About. halfof the RXD Events that took place in academic institutions proper fellinto the cat- egory of those that were recognizably mission oriented.Thus, 'the refereniced paragraph and Table XVI are coni~stent.

5.6 Contributing Scientific and Technical.Personnel

In tracing the hi-;tories of 'the utilizedinew science and technology,1,295 people were identifittd as hlaving mades.n. iat-otrbtin toRXD Events. Detailed personnel rdsu mds were obtained from 514., almost40 percent of them.- Analysis of those r6sumds provides the d~istributionspre! -nted In subsequent paragraphs.

Two relatively simple tests were applied-to the sample as a checkon its validity. In the first test, the distribution of r~sumds, interms of category of employing organization, was compared with the clis-tri-bution of Events that occurred within that organizational class. Theresults of this test are shown In Table-XVI.I.

Table XVII. DISTRIBUTION OF RtSUMfS AND EVENTSBY ORGANIZATION CLASS -

Orgianization class Events M% Rksumis M%

In-house laboratories 43.2 -35

Industry 43.6 53Universities 10.,6 12Other 1.6 0

The second test, involving a time distribution study, was to com-pare the percentage of the total number of Events that occurred in a~given year-with the percentage of the r~sumds that applied to Indivirittah-who had made their contributions in that year. The findings demonstr;7>ýa positive correlation at the 0.93 level.11 The rd-suinds seem to be ausefully valid sample. The slight bias In favor of Industrial per-formers wfill be considered later.

"lSpearman's coefficient of rank correlation:

r I- 6id 2 where: d = difference In ranks.nP:7l)n number of items ranked.

70

AA ki',ct fil I anld tinl ui~ cL 0 . r,i( ILr I' i of I ltI" lit. old I Ii t ,rI' I d t. i It iIh I L Ver~y 1111, OfI tile i ildi Vi tild , I 10S wh I L'At ie'c p itn'. ilidlte III ti( he Iti-, kniimvv tO Ii,11t' i1161.10 .1 lit i I iltu I L0nt r hut it'll 1 to cf~ti . 111 ,

c ontI- ri btit t ors arnd t he q ltjtncr) I 'i i L!tI I It C titId t .(i:CItII i L UIIIIII I Ue'111.1 i iV I I lit-iiati~ i rh uniii qt' t rpC of' Lite ' IIt il; tHIM t IOf C o'.- jut u)ppUr IInt t Yto t os I a numt' ibr III i~ypottIk-'? recor'tdbid qt product ivity andC tile cort'.e_Ltlt'nt..d' Of Ilatlage'r i~IpoIl cy aIc ti oil'. Stich cottip ri )r Sosan ilj Iy'~e art!preSen t-ed ini see t i OIis 5. 7 Ii routgh 5. 11

5.7 [ducat.ioni

Taole XVIII shows the, distributoio of academii c degree le~vel aniongthe identi fied contributors arid comparable data for the en Li re~ rat.ionalsc ient i fiýc and engi1neter ing colilwtiun ty.

Table XVIII. EDUCATIONAL LEVEL OF IDENTIFIED CONTRIBUTORS

_____________Source of Data* _____

flereeHINDSIGHT. NIORC TR NSaEC.

Ph.D. 10.5% 3.1% 'l.2% 3.8%

B.S. 57.0%- 34.6% 47.0% {3

Some college 6. 8%. 39..5% 4.6No college 3.2% 14.2% {46 3

Note: *Sources of data are as follow.s:-NORC: S. Warkov and J.. Marsh, The Education and Training of

America a. Sientists and &ngineera: IM6 (Chicago:National Opinion Research Cen'ter, University of Chicago,Octlober 1965), p. 17.

SRI: A;. SI p~ro,.R.P. Howell and JR. Tombaugh, Art Exploratory~ t~jof the Structure and Dynconics of thie R&D .tiduatry

.:(M nk1d PdrW, California: Stanford Research Institute,NSF *un6 19611), p. 3 1..

NS F: oiD es of Manpower in Sci~zce andTedainology (Washington,D.C.: National Science Foundation, NS 32,1963),

.EJC. "How Many, Eng inee rs " Engineering Manpow.er Bulletinj, No. 5,Engine.ers'. Joint Council ,New York, July 1966.

71

... ....

.- Th

:.z:.At the Ph.,D. level, data Irein the Stanford Research Instlitute (SRI)

-areprobably-the most useful for comparison purposes; that is, for Ph.0.,'i ~~~the 1O.5 -oercen~t ident ifiled through Project HINDSI[GHT ,sholu d be Lumipared :

with the 1.2 percent in the SRI column. The pharmaceutical*, petrochemi-o cal, biologlcal, anthropological and sociological iields, with their

relatively heavy concentration'of Ph.D.'s, markedly influence the datd.of the National Opinion Research Center (NORC) and the -National ScienceFoundation (NSF); this condition Is not significant in the SRI data be-cause that related primarily to the defense aerospace community. Theequipments chosen for the HINDSIGHT study mitigated against any strong"contribution from these Ph.D.-heavy fields, so it is reasonable to"exclude them entirely from this particular analysis.

At the M.S. level, the difference between the SRI jand NORC data isprobably not statistically significant,

5.7.1 Hypotheses: A comparison of the educational-level distribu-tions with those of any other source clearly suggests that the identifiedcontributors do not constitute a random sample of the general scientificand engineering community. Prevyalence of the Ph.D. degree Is more than

t •-eight times greater than would bd expected; the M.S. degree, over threetimes greater. -This observation might be explained, wholly or iN part,by one or more of several hypotheses. Perhaps-

naturead(1) -there is a positive correlation between the technical -

nature and the'amount of formal education to which an Individual Is ex-posed and his propensity for producing knowledge that is useful indefense weaponry; - -_

(2) the personal qualities that drive an individual to seek- progressively higher academic degrees are quite similar to those requl-

site to successful performance as a contributor of science and technol- I14 .~ogy useful to Defense; or

(3) the prestige of the higher educational level places theIndividual In a position affording greater opportunity to perform.

"The principal findings of Project HINDSIGHT (as presented in section.4) develop a strong case for the Do's support of directed research in

"science and technology. They do not-perhaps because of the more tenuoust'3" information flow and communication path between the universities and

Defense engineering-make as strong an argument for continued DoD supportof the less well-directed scientific Investigations In the academic en--vlrorment. There are good reasons for the DoD's support of universityresearch, however, and a quantitative demonstration of their validity is"possible.

An important part of the argument resides in the demonstration thatthe first of the foregoing three hypotheses is clearly valid and betterexplains the correlation between degree level and productivity. The

"72

~3, {

coni.eque,,ce of DD inAluence o00 the choice of scientific area of concen-tration it n the support-d tiniversi i(.i is % riman fested by tfe re I at i vel ygreater portion of those inst it ut ions' graduates who everntually make

..- ilitari ly significant contributions toscience and technology.

5.7.2 Sources of Education: Among the identified performers, 98!•r percent ot the degrees were awarded by colleges or universities In the

United States. The remaining 2 percent, awarded by foreign Institutions,were almost exclusively Bachelor's degrees, B.S. or B.A. In 1959, ap-proximately the year of the median Event conside-'r!d in Project HINDSIGHT,the university-trained scientific and engineering-manpower of the UnitedStates was slightly in excess of 1.09 million people.12 Of these, aboutS20,000 were immigrants who had received their education abroad. 1 3 Thus,4 .- on. the assumption that-there is no correlation between source of thedegree and propensity for productivity In defense-oriented science andtechnology, it would be expected that about 2 percent of the identifiedperformers would have received their degrees at foreign universities-precisely the finding of this study insofar as the B.S./B.A. level isconcerned.

The B.S. degrees awarded in the United States came from more than70 colleges and universities. The distribution among these many insti-tutions appears to correlate solely-and then, very crudely-wlth thei.,.•:r...•size of the university and its proximity to the industrial organizations

k-. F-and Government laboratories that contributed large numbers of Events.Both observations tend to support the hypothesis that prestige, or someunidentified personal characteristics, account for the disproportionately

'?high percentage of degreed individuals among the Event performers-atleast, at the B.S./B.A. level.

Table XIX lists fields of technical activity1 4 represented by theIdentified FAD Events and shows the relative distribution of B.S./B.A.fields among the performers in each technical area. The data in thistable clearly demonstrate a very markedcorrelation between the specificarea in which the person was educated and the one in which he usefullyperforms. Even though degree holders In technical areas other than thoseclosely.;ass.ociated with a given Event had essentially the same totalexposure to formal education-and presumably the same amount of degree-conveyed prestige-as individuals whose education was within the closelyassociated degree area, relatively few "cross-field" accomplishments areseen. This observation Is offered in partial support of the first

12 "Resources of Scientific and Technical Personnel In the OECD Area,1963," Manpower Statistice, 1954-1964 (Paris: Organization for Economic"+ '+•".•i~i •Cooperat ion and Development:, 1965)..:

13scientifie Manpower from Abroad (Washington, D.C.: National ScienceFoundation, NSF 62-24, 1962).

A 114CCSATIT Subjec~t Category Liat (DoD-Extended), op.cit.

.. 73

M

_4li-: .-

a)jiikkl tvopuNa a ,, ,, ,, ,

Af i bautb li I a a aa aaa a a a aa a a

4.4 S~ tt dLMIN 3aaanaavba.aaataa

-----------

JJutu s~flILtiu i: .a .a . a

00Z

4~~ -4-5,a

t t t 1 1,1 1

0 ~ ~ ~ ~ ~ t C.v- *,i~~t~6ue LugJpL a-aaa~ - aa ~ t - t a a a0 a 14 a 1

I I-~~~I~~4c Cc~P~ Ita~ a- a 1 1 a 1-a- a 0 aaa~~0s a a a aW a a a a aa~~~~ aa, a a a! a- a .a I a aaa cC a a 0 a 0, a ý a a a V a a a4* aU a- a a

a a. a a) a a Oaa a a a a aala ~ a, aý a; Wa a a0 a- a- a a a aip sa a a- a r- aaa a ba a a aaaa

a. a_ aa a_%. a.' 1- 0 a6 a a Wc a a~ > m ZZ a-aIF o t aa a a-a- a a a11- a .Ca aaa a='a a aaaa aa a a a a a ~ saa 4.a itaaaas ~ a a ce xaal -C a I, a a2a7 a aa C.Ž aaaoaý F

7alaa 4-a aa a aa a -,Ca ..-

........-

.1 IP LII I )I .1

•-.a u lAbtL.L,', i t)

AJJkIIIU M)LI'.) .I j A u h

-.Ji J L&LhUJItLPl 1 S t i t AVOI I MN

.. .. . . . . .-

s--. - -Q --U -- -- - -

"" 0

SE -

. S IF L. tiu Si•. 5S n LICES , ,i , ,

.- -- - -.- -.. .-..-. .- -. .-. .- -. .-. .- -.- - --. .-. .- -.

S... . d•iaoutbua Letilstrpul

•0

In. .aU u SP:) LSa L3

n I II420

a , u 4 c i m +' I Z S , , S .- C

:0 w it i I 114 .v II

C i l,, 5 i 5 4' 1 * w ii Ult

g,,l t t C i C U t Ll

.-0 .ItS I�. '�t"'. ILI S" Ili tl i l -

I..~• r_ r. a) 4) c

- do l 4j -S t

Elm. S r L I *z5i c 1g.,,

-.1.Z t ag

Stil t- Ill t 5.4) t41t S t

wt-t a 544 tu - S1. ~ w ittla)(D

CDWMi . VS -o). - i i tx r M: Xu gO 3

75.

lUtSt 5. L..441 5-+,I S -

AIJ IIU AV II J114 11 11 I

Iaufu **jiiu~a . . . . . . t . I

t a

I It 1 Is 1 1ip 1 -1 it Itt 1 1 I tI . 1 1 .- t-..

m 1 i 1 1 ~l l il 1 #SI 114

I-*t 1 1,*1 1 1 1 at I I4.

V1,i ii *iII 1t t ~ i t I

14 4* t I 1 4.4.1

I1 It ii I It11.1 1 .i.,i r- i

C~4t I I I4 11 I I i

C I C tIV I- t1 ii I i 0

wX Ul r I 'A h . I QU 0 In.-0IECI I -W

16 'W '~I It I io 4i *.i i04U

U -IC U V-).l 01 140 0 11 '41-4 .* 1 'A. .CV V 00.0 0- =4-'U U V1

UM.I 041 >J~ WW, 4 U4 0016 t.,1114l C 10t C C I 41C U

tj 4L.L== wm4.*.44iI S 11.- U-V)V

76

I' I

hypothesis stated above-that there is a positive correlation between thetechnical nature and anmunt of formal education to which an individual is

Sf• exposed and his propensi ty fOr*producing useful knowledge.

"Additional support for this hy'pothesis is 6btained by consideringthe sources of the M.S. and Ph.D. degrees held by the Identified per-formers. Al though the Department of Defense has supported research Inabout 200 colleges and universities throughout the.lUnited States, thelarger share cf this support has gone to a relatively small group of them.If the research efforts supported by the DoD were closely related to thetechnical areas of defense weaponry, and If there is a correlation be-tween nature and amount of formal education and propensity for producing

.• useful knowledge, there should be a cor'relation between the amount ofDoD funding support recelvedand the number of,identified performersgraduated. That Is, the universities heavily involved in defe se-oriented research should be producing a relatively greater number ofscientists and engineers with a'defense-technology orientation than thoseconcentrating their research activities more heavily elsewhere.

Considering first the identified performers who have been awarded aPh.D.: The top.23-universities, in terms of Support reLeived from the

-. DoD for research, awarded approximately-25 percentl5 of the Ph.D.'-s inthe physical sciences and engineering during the period of primary in-

•O .terest to the.HINDSIGHT study., Slightly in excess of. 50 percent of the"identified performers, or a factor of more than 2:1 over random expec-tation, received their highest:level degrees from those universities.

At the M.S. level, a similar situation exists. In this_ case, -thesame universities awarded about .11 percent 16 of the degrees in the United

• ,•States, but 46 percent of the identified performers received their de-grees from them.

S•As these are the larger.and generally: the better known of the.-American universities, they undoubtedly have a strong:appeal for thebetter students. It is quite likely, therefore, that the findings Justdescribed are influenced by factors other than simply the relative

Samount of research supported by the DoD. In particular, the phenomenon,"quality seeks .quality,.., can be suspected of affecting the-entering s.tu-dent. Nevertheless, there is such a marked. correlation between the-levelof DoD-supported research and the relative production of scientists andengineers useful to the DoD that the hypothesis of interest certainlyappears to be supported; clearly, thenature of formal education Isimportant.

1 5Privately reque.sted data from the National Science Foundation.

1 6Ibid.

77

5./.3 Consequences of ID6) Support of University Re e'arch: At this

I point in the consideration of the three hypotheses stated in secItion5.7.1, another matter of interest should be noted. The introductiont of

. this report (section 1) offered a hypothesis based on comine-nts by criticsof -certain DoD policies:

The support of scientififc research In the universities, by theDepartment of Defense and at current funding levels, signifi-

A- cantly affects the availability of the best faculty tVlent toteach students.

if'this hypothesis were substantially true, it would be expected thatS graduates of the less heavily supported universities would benefit-that,.- because the university devoted its better faculty to teaching them, the

S graduates should 'have acquired a better education and, everythinq elsebeing equal, should show up more frequently in studies such as this.Clearly, such was not the case during the time frame in which the iden-V. -tified performers received their formal education. The universitiesthat were most heavily engaged in DoD-supported research also produceda disproportionately high share of the scientists and engineers who wereimportant to the Dol).

'Further, inasmuch as the level of DoD funding in support of uni-versity research has not grown during the past few years as rapidly aseither (1) total Federal' support of research in the universities or(2) graduate education in the United States, it is most doubtful thatthe hypothesis would be any more valid today.

5.8 Military Service

In section 4, the argument was developed that the recognition of aneed is an important precursor to the production of utilized scientificor technical information. It is reasonable,-'then, to suspect that ex-posure to the military environment might serve to Introduce an Individualto the needs of the services and place him in a favored position to make

"a contribution. if-true,--this proposition could become of considerablevalue by-providing an additional selection criterion for employment inthe defense research'and development community.

In thesample of Identified performers, the roughly even divisionof those who had military experience and those who had none suggests.that, at least to a first-order approximation, exposure to the militaryenvironment may not be a significant factor. But the prospects of thisproposition are so-appealing that a more detailed examination of theavailable data is warranted (section 5.8.1); moreover, the proposition

'sheds further light on the hypotheses of section 5.7.1.

Uii addition, two possible tests for other differences between themilitary veteran and nonveteran populations involve relative educational"and age levels at the time of contribution (sections 5.8.2 and 5.8.3).

p .78

-SJ4

[7.

5.8.1 Nature arid Length of Service: If a correlation exists be-tween military servic and the probablity that the person Involved will

produce militarily useful technology, it Is obvious that the precise' nature of the military role he played would be significant. In that

"vein, 51 percent of those reporting previous military service claimedt that the nature of their assignments Was such as to contribute to theirfu~ture occupation. These qeneral statements are substantiated by the1high percentage of military job titles such as "radio and radar mainte-narite," "electronics technician," "ordnance technician," "aircraft :rnain-tenance," and the like among those who reported in that detail.

Other aspects of the veteran performers' association with the mili-

tary services lend further support to the argument that they are not anormal, or random, segment of the population. Based on populationýdis-tributions, approximately 10 to 12 percent could have been expected tobe commissio.ned officers; in fact, 40 percent were commissioned pr'ior toseparation front service. Of that group, 53 percent served during WorldWar Ii, and 47 percent, between 1947 and 1963. Similarly, of those whoserved as enlisted men, 56 percent were in the armed forces during World

.-Z. War I1, and-44 percent served during the period 1947-1963.

With the exceptlqn of a few people who apparently were called intomilitary service quite early In World War II and served until the end ofhostilities, the identified performers tended to have 2 or 3 years ofservice. Approximately 51. percent of those with military experience re-ported 2 years or less; 86 percent reported 3:years or less.

Thus, a "typical" veteran performer might be described as a youngenlisted technician or a technical of ficedr. He had bench-level experi-ence:with military equipment and a real opportunity tobecome feniliarwith special design or other requirements inposed on equipment by the-, military environment. This suggests that these veterans may well haveSgained a favored position, which constitutes an obvious advantage.

5.8.2 Educational Level--Veteran vs. Nonveteran: There are inter-esting differences in the distributions of graduate degrees held bynonveterans and veterans, the latter subdivided into two groups-corn-missioned officers and enlisted men. These distributions, shown In

.iTable XX, suggest that the commissioned veteran was more apt :to have ar higher degree than either the nonveteran or the enlisted veteraninthat'order. This ordering, although not denying that military servicehad put the individual in a favored position to perform usefully, tendqto support the second hypothesis of section 5.7.1' In fact, it appearsthat the hypothesis can be extended as follows: The personal qualitiesthat drive an individual to seek successively higher academic degreesand military rank are quite similar to those requisite to successfulperformance as a contributor of science or technology useful to defense.

During the information-collecting phase of Project HINDSIGHT, noattempt was made to ascertain the reasons for the differences betweenthe enlisted.veterans or all veterans and the nonveterans. However, aplausible explanation can be c:onjectured. First, It is quite likely

79K i

Fam

CD

4ad

9w-LI

800

hadt ai .1IM1U I -~i~ LIII d',t% COUI~t! and AI f& c-t .-ai bt- ittriL hIainged; thati t 111,1ý bi LJ- .i t'I hat at I. ,, 1 ,'o~ine of t hobLE viho tariwL d g radua te.

kleq rve-~. wL- v ~ , as btudent ~,hel Id by t he i r Selecct i ye Sc rv 1 Lv Boards ilii ad t Lif 1 r o a ~t t i unt i I a( Ie r r ad uat i un . By I tieli , mani y e fi e ti. iv I yavoidled ml i -ay se rv Icu be c au s ul' midr i t a Ist at us , relIadt ivelIy advarlcLed

aq rspec i al emp I oymeii t Second,* many of the pecr forme rs whu had the ir*advanced degrees bef'ore World4 War IfI were draf t deferred because of imn-

portant sci ent i Fic and technical assignmentb vital to the na t ionalI wa ref fort.

TablIe XX. D)EGREE DISTRIBUTIONS: VETERANS AND NONVETERANS

VeteransDegree Level Off icer Enlisted .All7 _ Nonveterans__

Postdoctoral 1.4% 1.0% 1.2% 0.4%Ph. D. 11.1 5i5 8.1 10.9M MS.-/M. A. Z5.0 19.6 . 21.5ý 22.9B.S./B.A. .55.5, 59.4 .57.3 56.e7..Some college 5,.6 10.3 . 8.6 5.7.Technical school 0 .1.4 1.0 1.7 .

High school only ý1.4 2.8 2.3 1 .7

It was noted in section 5.7 that there is a correlation betweenlevel of academic education and propensity for advancing the state of.the-art of military-science and technology in.a useful manner., Thearguments presented In the foregoing several paragraphs suggest .t~hat*practical experience with military equipments., or perhaps attendance atone of the military services' schools 'in pr'.paration for such an assign-ment,.17 *tends to give a measure of equalization. This explaniation 'wouldrationaliie the generally lower educational l~evel) of the enlisted vet-eran, who~is the one most likely to have taken a course at~ahighly

* .. specializeditechnical service school. Further, It supports the hypothe-*sis that a correlation. in fact exists between nature as well as level of

education and tendency to perrform usefully in Defense R&D.

5.8'.3 Age at Time of Contribution-Veteran vs. Nonveteran: AnAexamination of relative age distributions, veterans vs. nonveterans, atthe time a contributlon~was made produces inconclusi've rresults., Someveterans completed all *or part of their education before entering themilitary ser-vice. Others started undergraduate or graduate work afterleaving the service. Among all performers, as shown in Figure i5, there

Y ~was considerable spread in age at the time of contribution. The Inter-qua;:ile range of what approaches a Poisson distribution is (-)3 and

"1This information Was not solicited; because of the known tendencyof the services to give special schooling to technicians, however, thepossible imnpact of this factor cannot be ignored.

81

1. y vit >.I pte.I1,1 i ! x iI~l I I d lit' I YItju i I ~ ýj Ii t

0 k ..4- I Iu if, 1) 'ilt I aqw (A Vt- v Ii all. ,u ai fudve I I-Ial.

Mob Mi I it

MAI ,i Il it y od ivnt ii . .1114 et VA illi LJ 'l ' It '.UInileI ;-;fi hi! dti!'AL.. i Ittj ill

I .if dill. et-te if fil1.. Ilnfit I v '.h Ii y",' the ptol't-f%'i 1)'

of Il ii hi V i dlalI I titn o%% ii' fill I fiv' itur".ly t iq zi d b ll ieif O'iii kit ic' i,

If,- ii t n I ort lIi otn,' I L% It ol I I lk 'I oh I hi VdI " wol,t en at I y'l w! Ill ) i-- en% i ia I. %

. A, ~~~t t'ehni Ial vept'r I i .v (st tit ill Z~9) -1Ile t- I I. Ilso "i U -iorga i /at i unaI"kobi Ility.'" a'ý. .1itrtl t-1.on citi, h le nature of' his~ win k wi thift If inj It0ort.1a i .at i on.* i . e. , Chaulge.; I) I h't- I hl~am thios'' fi o Ilvi Inq 'k, iel e i I ic Uirt i'i111 hni , I .'a'; ol iit orve., I (%ctl ion !' .'. 3)

Of th e',i' thu i' foruit' of mo initti ty only Interorgan izat ional n"tvelueniti!, klititil to have been studied to any- great depth. Shapi'ru and other!,invest igited tho movemnwrt- of several thousand sci nti~t~ and qrliifer it.,priia~illy of the acrspoce research and development conimnuni ty. " To al imi ted extent , the chau'ac ter Ist ics of the i nd ivi dualIs id ent IlI hdI hrotiqh Project HIlNDS IGHT can be compfar'ed with the SRI rindi nts.

III the followting sections (5.9.1 through 5.9.4~), HINDSIGHIT data arefurther analyzed to continue testing the hypotheses presented earl ier11nth to provide a base I ibe fu~r future studies of the behavior patternsof scientists and engineers.

5.9.1 Professional Mobility; 'This kind of mobility was definedin the foregoing paragraphs as the propensity of an individual forcrossing the loosely rLecognized boundaries of scientific disciplines orart:as of technology. A measure of the professional mobility of theidentified performers is obtained, through analysis of the matrix i

Table XIX, a comparison of the fields of technical activity representedby the RXD Events with the relative distributions of degree fields; atthe B.S./B.A. level among the performers associated with each, technicalact ivi ty.

as r n purposes of analysis, professional mobility Is further definedas ncopasingthe expectation, that a given individual will make a

significant scienti-fic or technical contributio n in a f ield other thanthat for which he was prepared by formal educaltion. That expectation,then, is measured simply,-as the-ratio between (a).the number of individ-u als who, because of degree field, might normally be expected to make noconttribution in a specific area and (b) the total number of individualsmaking contributions in that alrea.,

18ROI Studies Series (Menlo Park, California: Stanford Research

4 ~Institute, 1965-1966).

82

ti i:'?7

Iull'III the '.dIIit, raw dato t hI-t pruvi itded Table~. Xl X, t p rofes sional -,i -Ilhi I i-l" i uhx fur- pscrfurirrt~r' with B3.S. or- B.A. d~i-f'iCalculated tobt 0, 1 /. il I'l r t .11 crilationis Jul- er urrlii~mt'r. whose Iriqhce'st degree isM.S. .r M.A. rvesult i 0o ,)I I ~itr liy iiitlt'x of 11.-13%. For- fle I'l. D. level,all i.ndvx Lit 0..01f i.. to' ill. 1 .I.. xpecttud inur Iii t y at the. m. S./M. A.lL'vt' I i, me1rt. t han thlrrel t i ies' gredt~er that ýit il at thtt Ph.D. 1levelIIfuk 111101.11 l0111- HIM ir I 'Vlts1 cr t i lt thll B.S./tt.A. level.

The rank or-deuring of' this finding is not at all surprising. Ther*Ca 1i .at ion that 01v dat a suggest I haL only sonit' 4n percenit of the. Ph.D. 'scan be expected to make contributions outside their originally chosenfields, however, stijges ts that Ph.D. trainfing within U.S universi tiesis quite narrow.

Ilii part , perhaps, thib *i nding-thfe lack of' professional i mobilI Ityonl the part of Ph.D. 's-explains the earlier finding of a c.orrelationbote,.en noD-supported -research in universities, and the defense orient~a--tion of those universi ties' graduates. The degree fields shown as-theabscisso of Tabl e XIX are obviously quite broad. Within .thes.e fields,each Plh.D. candidate selects some subfield for specialization. If he isinfluenced in this choice by a faculty adviser engaged in boD-ýspons~credrescarch, if his research is oriented to military uses, and if hismobility (i.e., movement) from the subfield is little greater than it istrom the general field, it *is not surprising that he is found later tobe making a useful contribution to science or technology significant tothe DoD.

ýThese arguments further support the. first hypothesis of section5.7. 1 ,that' there Is a positive correlation between the technical natur~eand the amount of formal education to-whi~ch an individual Is exposed andhis propensity for producing know~ledge that is useful in defense weaponry.

Realizing that the faculty for an M.S. degree candidate is, to aconsiderable extent, made up of the-Ph.D. candidates just mentioned andtheir fa~culty idvisers, It Is almost to be expected that the M.S. candi-date would become "defense oriented."

Why the M.S. degree f rom a un~iversity where the DoD supports re-search is relatively so much more. prevalent: among the identified per-formers than the Ph.D.dAegree Is not readily apparent.. One possibilitythat cannot be discounted is that exposure to a defense-oriented facultyhas caused a greater than average percentage of graduates from theseuniversities to seek employment In the~defense~industries. This matteri's :researchable; such a project may already have been undertaken, but,if so, -the findings have not been widely circulated.

Anothor but more subjective measure of the lack of professionalmobi *ity among the ideritified performers may be seen by comiparing TableVill (a list. of thme scientific and technical areas Involved in RXD.Events) , in terms of. frequency of occurrence, with Table XXI, whichshows the distribution of degree fields by educational level. In

83

,iddition, Table XXI gives the current diktribution of the defen.ti: KLOwork force in the Los Angeles area as determined by the Stanford ResearchInsti tute.

¶ Table XXI. DISTRIBUTION OF DEGREE FIELDS BY EDUCATIONAL LMVLL

HINDSIGHT sample Los Angeles areaField B.S. M.S. Ph.D. (al de_•rs

Electrical engineering 30% 33% 19Z 23.4%Mechanical engineering 26 16 7 18.7"Physics 15 21 18 7.1Chemical engineering 7 5- 7 5.4 )'Chemistry 7 3 0 ( )Aeronautical engineering 6 5 5 12.4"Mathematics 3 12 14 7.5"Metallurgical engineering 1 2 0 0.9Civil engineering 1. 1 0 2.0Industrial engineering 1 1 .0 1.1Physical chemistry 0.5 1 7 NS§Nuclear engineering 0.5 0 0 NSNuclear physics 0.5 1 0 NSEducation 0.5 1 0 INSBiology 0.5 0 14 -0.5Organic chemistry- 0.5 1 2 ( )

Notes: Columns do not add to 100 percent because of roundhirg.S*( ).,�-- Included within chemistry.,§NS - Not shown separately. -

l With the primary exception of the :aeronautical engineers, the dis-tribution in the Los Angeles R&D complex is seen to correlate reasonablywith that of the HINDSIGHT-performers. The exception undoubtedly isoccasioned by the prevalence of the-aerospace Industry in that area.Again, as In Table VIlI, Table XXI offers some guidance on the relativenumbers of scientists and engineers (by discipline) that will be requiredIn the'near future, assuming no major change In the nature of the equip-ments to be developed for the mil itary services..

5.9.2 Interorganizational Mobility: It was noted previously that"the biographical sketches of the contributors Identified through ProjectHINDSIGHT were collected during 1966 and were current as of that date. -

"Thus,.it is possible to examine the Interorganizational mobility of

9IA. Shapero, R.P. Howell and J.R. Tombaugh, :The Structure- andDyn,,ics of the Defense R&D Industry: The Los Angeles and Roston"Complexes (Menlo Park, California: Stanford Research Institute, November1965), p. 24.

84

*thost. peoplet ove~r d l ong period ot' t Iiie- in somfe cases as miuchi as; 4oyears. III :Figure 16, Lhe sample of pe rform'ers is exami ned wi ti retapectto total years of piofessional clivi I an experience Vs. frequentcy o foc c ur-re ncc. Half of the sdinple had between 13 and 24 year% of experi -ence. The median length of employment was shown to be in e~xcess of" 16years.

Shapro nd thes stdie th eninee/scentst.work force of thedefnseR&Dindstr inthe Los Angeles and Boston areas.~" Based on

abot hlf ~ht fundforHIDSIHT &Dpeople. The distribution aboutthe median for the R&D population in.the Los Angeles' and Boston com-

-~plexes forms a pattern very similar to that shown in Figure 16, except,of, course, that i t i s marekedlIy d~isplIaced. toward the or ig in...I. Des pi1te theobvious Co'niiderable difference in averige total length of pgrofessionaiemployment, the HINDSIGHT performers appear to have made fewetr totalmoves. Comparative data are shown in Table XXII.

Tabl e XXII. INTERORGANIZATIONAL MOBILITY

No. ofPrevious SRI HINDSIGHT

FEmployers Sample .Sample

0 24.7% 29.2%

1 21.8 26.19 0~.20

101 1r4or-037

Unobe3 the. vale ofmn3niiul dniidtruj hHIDIH s9uie ha&ogbe.rcgie4ythi~epci eepoes

sucssu perfmormer is3 nospoet cag o ahsaeaepo

hesoa associate..ndvidal

Undobtedly t pp idn2fe andug 23.

HIDSGH sudeshd on benreoniedbythirirseciv ep85esIt-ranycnb xece htwaee preito hs mlyr

exrsepu htve aifcinteidiiulsdrvdfoknwn4htterwr a en uecnrbtd'oterltv

qps G i'

Zcc

Lai

00

-- /-0

tDL

I -4

U- 1^0

8 06

A somewhat di fferen cause-and-effect relationship may be inferredf "romr nthib dppdrently low mobility, along with tl data on age at thet : ' time of. ontri bu tbtion (see Figure 15). ,Together, they suggest that theperson'had been witl hiis employer -for some time before.making the notedcontribution This could mean that, because of .his long service With

. the organization, he was in afavorable position to-.have his ideasrecognized and exploited. The inference would be entirely consistent-with the finding (section 4.3.7) that: A high combined inventiveness,or ingenuity, and'utilizatlon rate are dependent upon the time and spacecoexistence of four primaryfactors-the recognition of need, a sourceof ideas in the form of an educated talent pool *capital resources andan adequate communication path to potential users.

The general educational level of thecontributor already has beenestablished 'Add to this the fact that, through long years of associa-tion with his company, he had acquired specialized experience in itstechnical affairs and.had developed the ability to interpr.et Its needsand problems. Then, onlythe resources that one could expect would bemade available to a trusted employee are necessary to complete the fourfactors named in the. f in~ding.

On balance, it appears logical to conclude that it is the indi-vidual's low itnterorganizational mobillty that leads to succes~sfulperformance rather than his perforimance that leads to job stability.

5.9.3 Intraorganizational Mobility: The HINDSIGHT contributorsshow as little propensity for moving within a company as they do fornioving between organizations. -The Information they submitted suggeststhat, on the average, each person had held two different positionswithin his organization. In most cases, he changed from a job callingprimarily for activity In a given scientific or technical field to anassignment chiefly Involving management responsibility in the same tech-nical area or one of somewhat larger scope.

This finding appears to be entirely consistent with-and possibly"helps to explain-the previously mentioned finding that there Is a veryhigh degree of professional stability among the identified contributors(see section 5.9-1). To some extent, this stability may be encouragedby the fact that senior management may be reluctant to make assignmentchanges that.would force the employee into "professional mobility," i.e.,into another scientific or'technical field.. Of course, it may be that

, management merely accedes to the emp:loyeel5 Insistence that he be al-lowed to remain within his chosen technical area of work.

This is a researchable question, but it was not addressed in theHINDSIGHT-Task I study. A more thorough understanding of the matter

would be useful, because it may be that management has inadvertentlyadopted a policy that mitigates against the fullest exploitation of itsmost creative employees.

87

1.2"

5.9.4 Education and ReKponsibility: The third hyputhitsIs in sec-

t tIon 5.7.1 suggested that the correlation between level of education and

propensity for making uti l ized coit ribut ions might merely be bas!ed on

.A&-,• the fact that the presti.ge associated with the. higher academic, degree

held by the Individual placed hiim in a favored position to perform suc-cessful ly.

"A partial Lest of the hypothesis can be made by .comparing educa-

tional achievement with other measures of success within the profession-

al society. Specifically, degree level can be compared with-the Indi-

vidual's hierarchical rank-within the organization. To facilitate this

"test, six categories, or "responsibility/authority levels," were estab-

lished, as shown In Table XXI,1 and the sample of identified contribu-

tors was separated, by degree level, into those classifications .with

the results given i-n Table XXIV.

":1 . Table XXIII. RESPONSIBILITY/AUTHORITY LEVELS

i'• Level Assignment ..

Division director Supervision of several branches related bybusiness or object interests.

-. Branch director Supervision of several groups related by. - • • ~ ~discipline . . . • : •

a Project manager Supervision or coordination of two or moregroups related by project.

Group manager Supervision of one or more working-levelpeople.'

Ž" Professional Degreed scientist or engineer at work level."Semiprofessional TeChnician.

it is obvious that, in absolute numbers, the top four responsibili-

ty/authority levels are dominated by people whose highest degree was at

the B.S./BA. level. This finding is partially a consequence of the

fact that the largest percentage of the identified contributors is in.

the Bachelor's degree group. The distributions of the managerial segment

of Table XXIV can be adjusted to compensate for unequal degree-level

distributions among all performers (Table XVIII), with the results

shown in Table XXV.

88

tr

ismIn•muu unnnn- mu

Ta h1)1t XXIV. LINNE. 1 Gi ITEE) Dl STR IB 11 ION: MGM CIl. VS. RH[SPONS IBIIL. ITY LEVLL

- [duc~at-ion -1

Soiie -. HighRqRLiaSibi yýLevel Ph. V. M. S. 11 US./B. A.ý cc] lqqe school I

Division director 26%. 20Z 51% 3Branch director 13 22 35?Project manager.. 6 ?25 61 6 2%Group manager .29 58 4 2Profess'ional M 2 61 6 2

Seirfesoa -10 41) 45.

Note: Horizontal rows add to 100 percent.

Tab e XXV. WEIGHTED DISTRIBUTION: DEGREE VS.. RESPONSIBILITY LEVEL.

______________ Educationhisome HighResponsibility 'level Ph.D. .M.S. .B.S./BM. college, school

Division director 43% . 33% 16% 8Branch director 34 . 28 26 12-Project'mnae 13 262 21 15%..Group manager .15 32 *24 14 15

At the two. highest levels of authority, degree level1 and responsi-bility are seen to be positively~corr0.lated and monotonically decreasing.Recognizing that the degree fields of these indivi'duals were science andengineering rather than managemient~ this finding does suggest *that theprestige conferred by the higher degree has an influence in achievingpromotion. Or, as in the case of recople with military experience, Itmay be that whatever personal characteristics drive one to seek higherdegrees have much in common with ýthose that impel one to seek positionsof responsibility.

KThe dist~ributions at the' levels of project and group isanagers arenot at all inconsistent with either or' both interpretations,. The~dataon responsibi~lty/authorlty levels relate to the; contributors' currentpositions, not their positions at the time of the 'RXD Event involved.

.-J Further, the length of-time that most of those individuals had been.Afiemployed suggests that, in the preponderance of cases, they were at or

near the highest positions of authority to be expected. Thus, Table XXVdemonstrates that there is a weeding-out process and that the personhokC ng the higher degrees can. expect commensurate rewards.

It is clear, however, that this analys~is-intended to test thecorrelation between degree-conf erred prestige and propensity for making

* militarily useful contributions to science and technology-results i-nequivocr-I findings. Prestige may be a factor, but it does not appear tobe truty sigr. [cant. 8

Further analysis of the biogrziphical sketche-. suggests that inter-ofganizational stability ,tlso cur~relates positively with respon±.ibility/author•tv level--,L least, among thin identified successful performier,.Table XXVI displays the relationship between number of previous employersand hierarchical level of responisibility achieved by 1966.. Only manage-ment levels are considered.

"Table XXVI. RESPONSIBILITY/AUTHORITY LEVEL VS.NUMBER OF PREVIOUS EMPLOYERS

S" Responsibility/ No. of Previous ripyers

Authority Level 0 1 2 3 4 more

Oivision director 28% 19% 28% 18% 4% 3%Branch director 32 21 22 6 10 9Project manager 33 26 17 10 .9 3Group manager 34 33 11 14 5 3

Note: Percentile by row.

Comparison of the information in Tables XXV and XXVI suggests that,at the division-director level, some tradeoff is accepted between the ,Ph.D. degree and number of previous. employers. That is, at this highestlevel the Ph.D degree Is somewhat more important than interorganization-al stability as a criterion for promotion. This is additional evidence4. that prestige associated with degree level correlates with the attainmentof positions of higher responsibility and authority, It is noted, how- Iever, that this particular finding Is not necessarily relevant to thehypothetis concerning nature and level of education *and propensity for

r-. useful .contribution.

•.l 5.10 Marital Status

for the current engineering and scientific work forces 'of the Bostonand Los Angeles R&D complexes, Shapero and others reported that 70.9 per-cent-of the males were married, 27.3 percent were single, and 1,8 percentwere divorced or separated at the time they were hired. 2 1 The somewhathigher average age of the HINDSIGHT contributors at the time their bio-graphical sketches were collected (43 vs. 36 years) precludes a usufulcomparison. Nevertheless, the differences are great enough to warrantreporting. The distribution among the HINDSIGHT contributors was foundto be:

Married ------ -------94.0% Widowed- ----- 0.4%Single -----------------. 9% Ever divorced--4.0hDivorced or separated-- 0.7%

"A� "Ibid., p. 26.

ot

I t i i pi mar i Iy Itlie d i ftIrenFue i n thle ''divoi teu or- separated''If (Je'e., hSp i (.0 110 re lo dL 1 e 1 y sma I Ipe rcentiuje in e ithle r ca!,o, thatjma~y. bk, ITLip, t .n1teRe'. t i nyj The, repea tedl1y notvd tendency toward s LabiI It y

i i tt.' nuS of 'larts (if profe~ssioita I intuest , in t ermns of einpl oyerý , i rterms of pos itions wi thin an o ry i ot (flis utle*cv(endbythapparently qreat("r-marital stability of tile successful contributor.

5.11 . Continued Po'~tgr~aduIate Educat ion

I n add 1t icril toformalI educa tiui oleI ad ing to the award of a degree,many of' the con tributr re Lred attending one or more postgraduatecourses. Well over 50 percent of the individuals partLicipated in short.courses or lectures. series. sponsored by their- employers. The avai lableinforniation onl these courses, w hich v ar ie d in length between a few1hours and a 'few weeks, is inadequate as a basis for analysis of thequality or significance. of the instruction. A somewhat smaller groupreported that they had taken formal courses offered'by recognized uni-vorsitles ei~ther on or off.campus. Table XXVII presents a distributionof the latter information in te rms of rthe highest degree' held by thereporting-person and by the general nature of the cours .e.'

Table XXVII. POSTGRADUATE EDUCATION

Type-of Course Ph.D. M.S. Be.S

Technical. 6% 14% 10%Admi n!istrati ve 0 4% 1

In some cases,,th e individual had delayed..taking a formal coursefor, as much as 20,years after receiving his. lost degree. The, average'delay appears to be abou~t '7 years, and the median, about 5 years.-

No.-comprehensive information concerning the practices of the'defense-oriented R&D cormmunity. s' knowri to 'be avai lable. ,There. are

soedaeo individual rorgn z on n on the entire national scien-

tific and engineering community', howeve~r,' that offer a' limited basis 'for comparison..

*The National Science Foundation22 estimates that about. 5'percent ofA." the country's scientists, and engineers are currently taking part-time

university work. From thte NSF datal i~t Is not possible to determine thetotal percentage that conttinue with postgraduate education at one timeor another over the years.

22Zola Bronson, National Scie~nce Foundation, private~corres~pondence,August 1967.

91

At th,. individual urqani/atitonal level, I.l. iMtRuhlin report-. tlIt iby 1967 fl1owe than 90 percent of thie profe'.siunals at the Langley Research

i' Center, National Aeronautics-and Space Administration, had taken at leastttone formal university course under the auspices of NASA's continuing edu-cation ,program, with approximately one-third `6f the staff participatingin any recent year. At the Army's hlarry Diamond Laboratories, 85 percent

- ,of the professional staff completed courses of at least one semester's-A duration during the period 1964-i967.'I-

By any of these measures, the Identified contributors to RXD Eventsare markedly In the minority.

"One can speculate, of course, that the con tributors are leaders intheir fields, and therefore are more likely to develop information to be

. taught their peers than' to learn i t' in a classroom. Or it could be thatthe contributors tended to travel so much in the course of their workthat taking even part-time classes was impracticable. Whatever thereason, it is apparent that these valuable, creative people :were notgiven to exploiting the opportuni ties for continued'postgraduate educa-tion that are offered by most of the larger universities.-

5.12 Summary and Discussion

These additional findings combine to present extraordinarily con-sistent patterns. Most of the scientific and technical Ingenuity andinnovation leading to modern weapon systems is found .in a very conser-vative, almost routine environment. For example, nearly half of theadvances came about as a consequence of what the performers consideredto be routine assignments (Table X). The persons credited with theadvances give evidence of extreme conservatism. In comparison with

- •others in their professional society, they tend more to remain in thescientific or technical areas in which they were educated, to stay onthe same Job with the same employer, and to be satisfied with the levelof education attained before they first sought -professional employment;they even tend to have more stable family lives. The chief possibledeparture from this otherwise consistent pattern is their apparent will-ingness to move from their. chosen. technlcal fields into managementpositions when they were. relatively young.

To a considerable degree, this may explain why comparatively few of",t- the technological Events were found to be based on the results of recentscientific advances. For Is it likely'that-individuals with-such stableand conservative traits would delve Into what is, to them, 'the unknown-" '"fOr the solution toea problem? 'Certainly not, If such a digression

E c231n.M. Rubin and H.G. Morgan, A Procj-ctive Study Toiaads Con•.tinring

EducaioOn (Boston: Sloan School of Management, Massachusetts Institute ofof Technology, 1967).

2 4Maurice Apstein, Harry Diamond Laboratories, private correspon-dence, August 1967.

42 92

t i'4g

coulId be 41 voidtd. I f the characlvfi~tics (if people maikhin ,19nificant!

techno logic i) con uiibut.ituios In ai~reas other than weaponry are equalIly

contcervat ive , there -~ iIi Ltile wonider that new scient ific findings move

so ; lowly into vveiyd efy ice

75Jacob Schimookler, Xnvention rand'Economie! Growth (Cambridge:

Harvard Univer'sity Press, i966).

93

. Preceding Page Blank

6. RESEARCH ENVIRONMENT FOR IIIGH PAYOFF

6.1 RDT&E Envi ronnments

Management hat tended to create different environients for funding"control and the decision process in the several phases of the RDTIEcycle. In addition, it is alleged that significant intellectual andsociological differences may be found between the classic organizational"structures identified as "Government in-house," 'university-associated"and "industrial" laboratories. Presumably some one combination of thefactors making up the various envi ronments should be more effective than"others for fostering creativity and the utilization of its products.-

One of the prinary stated purposes of Project HINDSIGHT was toidentify management factors that constructively affect the productivity"of a scientific or technical research program. It Is emphasized thatthe factors sought were those associated with the combination of creativ-Ity'and use of results. Quite possibly, the optimum environment repre-sents a compromise; it may not be the best situation for scientificcreativity alone, or the best for innovation.

An i rmediate observation based on one of the HINDSIGHT findings(see section 4..3.1) is that the broad categorical differences between"Government, university and Industrial laboratories do not markedly in-"fluence productivity In terms of the emphasized combination-creativityand result usage. Similarly, the study shows that productivity is not

A• restricted to the environments associated with the R&D categories of"either research or exploratory development.

"Based on the 686 RXD Events on which pertinent data are available,the percentage distributions among the Defense R&d categories were foundto be:

Research-------- -------- 9%Exploratory development--28-Advanced development-...O%Engineering devetopment--22;Management and support--->l1

A considerable number of Events occurred before 7 June 1962, thedate on which these categories were formally established. Those Eventswere categorized, insofar as possible, through analysis of. the c rcum-stances in which they came about. When there was significant doubt,the. Event was ignored in calculating percentage distributions. Thus,"the d:stributlon is only roughly descriptive of the situation and Is ofinterest primarily because the peaking occurs in the advanced develop-ment category. Obviously, creativity and the Imaginative applicationof scientific principles can be found in work relating to every phase of

95

iho' RM*it. E y L I~ v P1 ot!oCI i vi ty s' ý.sitjni I i, ail( ly Ilifv h tel L cd i 111t.-i by I ,,r

Col id ti rd hesX re or bly ciii I t Nel t . of' f",Ituir. thilt opp ly wi huti d

Cat ytI (9iy

1h i s- I tidy wa p, ut dLvLp Lntigh Jto u ellsv dill trcn t s.et' (if f aclorji f Inq IJ n. thLey do ex is t. Subseqtienl jdrdyraijhit. Cortsi de~r t hie f infd-

iqs. wi th. re y ird to apparent I Y, sic ~iii ujictvi env ironinmen tal factors~.. 1tcs~eviea af c I iss ' s r si at 1Lb IF rom und i rvc. t ed st. i elitj 'ic, rcse6 rch I.h rouugh

Arst Lit,1ý1 inl tL1 lnoloqy , ae t. r eated Ii Irit un, and the fi tidings, are di s-cussed in a elation to ie rtIin.itl maiauqrriient l. .c tors tikit scent to becohntollable by pol icy.

Based on the methodology employedl in Project HINDSIGHT, onlysuccesses were st~udied. It. co.nnot be concluded, therefore, that this*'must effective" environmetnt wi ),.I ensure success,;, fai lures have hsappcrene~aod probablIy will continue to occur,... Because other patterns of manage-inint have failed to exhibit as high a etlzto ae hoee,itrusIbe. assumed that their efficacy was loYwer and their failure rate, at leastsl ightly higher.

To this onPlayoff has been. considered only as the recognized

us e of results proceeding, froam, the research effort. on, a military system.'IOther forms.,of payoiff'were observed in the course of Project HINDSIGHT,*but the establishment of their relative value to *a ntisslon-orientedagency, such as the DoD. is .beyon~d the scope of this study. It Is clear,however, that they do exist and, by and large,. tend to be associatedwith generally~unidirected research in) science. (See 'section ?. on HIND-SIGHT u~bjectives , st rategy. and me thodolIogy. .

6.2 Research, in.Science* 6:.2.1 Undirected Basic Research: There is very little evidence *

*that ideas spawned during recent basic research, particularly in'the so-called undirected basic research In science, as opposed to directedresearch -(Tables Ill aind IV), manifest themselves In improved weaponsystems. In the fields of statistical analys-Is, other aspects of infor-mation theory, nuclear physics and 'polymer chemistry, A'very few exam.-pies. have beein found in-Whith payoff was identified in a system.

Nevertheless, it has been seen that, wherever there was a payoff ofbasic scientific research in termis of use in systems, its value wascomparatively~great. Exampl~es are found in eill solid-state electronic* sstesnuclear weapons and power supplies, Information processing in

computers and radars, and the application of statistical sampling inquality-control techniques.

The greater payoff of undirected basic research in science ldcnti-fled in HINDSIGHT studies has been' obseirved in forms other than-directweapon-system application. In terms of numbers, the gain has been in

96

IIthe t ra in inq o f scientists and engineers who performed the direct ed rv-search in basic science or in technology that led to the new weaponsys tems . The technique of investigation,,the sd--cal led'sci'entific ap-'- proach, that has been developed and propagated by the universities ha,,had a tremendous observable impact.

,ection 5.7 demonstrated how valuable university-trainedscientistshave been to the Department of Defense-in particular, those trained In

Suniversi ties that had defense-oriented research programs. ' When the DoD'srecognized dependence on schooled scientists and engineers is considered,

Salong with the fact that in a typical year (PY 1965, for example) approx-imately '6 percent 26 of the nation's scientific and engineering talent is

Rengaged work that is directly or indirectly Defense supported, the'magnitude of the payoff may be appreciated.

The second very useful product of undirected basic research is theinstrua~entation developed by thescientists In their quest for knowledge.If there Is a characteristic difference between undirected and.directedresearch in science, it Is probably that the former examines a ll- possibleaspects of a phenomenon in detail, while the latter tends to be morerestricted in scope, focusing on the apparently most pertinent aspects.In their desire to observe in totality, the performers of undirectedb asic research have developed- Instrumentation that is exotic and precise.The -resulting devices were Identifled through Project HINDSIGHT as they"became :the tools of the applied scientist and-the engineer.

The nature of a.thlrd-payoff of undirected research is suspectedbut has not yet ben well established. 'Often the candidate for a

-Master's degree in science or for a Ph.D. bases his-thesis or disserta-tion on research into a particular characteristic of a device or mate-rial. If the results of this research in any significant amount arecollated by the publishers of the commonly used engineering handbooks,this third form of payoff is established. This matter is being investi-gated as part of Project HINDSIGHT's more detailed studles of banagementfactors.

2 7

6.2.2 Directed Basic Research: No sharp ]in, of demarcation isfound between theoretical, undirected research In science and its di-rccted counterpart. Instead there is an -almost cont I nuous spect rum,trom the case In which the -scientist.undertakes a particular line of"nvestigation r"because It looked interesting" to a ventute that Is

totally- responsive to the request of an outsider.

2 6Estimated from the ratio of Defense spending to total national,expendi tures for R&D. Reviewe of, Data on Science Resources (Washington,D.C.: Na-lonol Science Foundstion, NSF 65-Il, May 1965), 1, 4.

S!7Ro,;ert C. Mills, Liaison Activities at IMP Interfaces-A Model;Some anpirical Resulte and Design Considerations for Further Study(Evanston, )Iilnols; Northwestern University, June 1967,1.

97

S il i• i ~

Mos.t Of tile %Lielltf fL research 'sup ojrte( I'dby he Dot) is Ut.: 1itfby seine rat ionale anid is typically refor~red to as ''r levijit ."The amll; I:ysis of Project HIINDSIGHIT data on) thle use of results of directe~d sCi..n-tifi resarchoffes spport for the c~uicept o~f re ltvaricy and sgetthat a rigorous defi nit iconof relevance' is essential.. Where results,Wert: applied, thle research task tended to be relevant. to a very specifictechnology already being used in a cruder form~. I n a' sunse, recentscientific research' that hjs been found significant to improved weapon

ssestook the frofa "mnop-up" effort. Engineers were employingatechnology with sme de~gree of success but didn't understand i r d etailhow or why it w6rked. Scientific analysis led t'o a better understandnand even tualIly to greater~ exploitation of the technology.Exainiples of the return on investment 'nsi ntfmo-preac

were found in-every weapon system studied. Problems of ext~racting* information from contemporary radars led to filter-theory research.4 which,Jin turn, led to far more capable. radars. .Propeller-noise prob-lerns led to cavitation-theory studies, that. generated the knowlIedgeneeded for designing less noisy propellers. fo *r torpedoes. Problems ýof1ncombustion instabilit in rocker and turbine engines furnished* the....rationale for pertinent scientific research.,

Relevant research probably results in a high p ayoff in utilizationfor two interrelated reasons: (I) The scientific research was known tobe in a useful area suggested by a problem i~n the associated technology;and (2.) *a ready market-for any new knowledge generated was known toexist1 for in-advertisihg theirlproblem the engineers i.,denti~fid :them-selves as potent~ial useris. Profitable directed research, as~describedin 'the precedinag examples-, requires a fa'iri.y direct line of~communica-.tion between the engineer -and the scientist. .-

6.2.3 Research ih Materials: Another clas.s of research in scienceobserved in the Project HINDSIGHT studies has had an equal or greaterpayoff. -el :vehalf the identified technological advances., clearlyoccurred in the absence of new scientific ideas. -Almost all,.Ihowever,depended upon the existence of new information on characteristics ofmaterials or'thle operation of devices such as the transistor. The classof reearchthat l~ea'ds to new materials., *to measurementsofths aerials' characteristics,. and eventually to the publication of pertinent4 data has been-of pelrnaq~ value.

From the pragmatic; view of a mission-or iented agency, faIIrlystringent criteria for the class of research lIeading. to reference -textsand handbooks can be established. The producers of the documented in-formation, or their managers, must remain aware of the technical prob-

Sectio 5.1.2.2 discussedi the lmaortanr-e of identifying, publishing

and dissemuinating information regarding thle characteristics of new

98

ORA

Alaterial. In this particulr exoaxp iple, the. informaation i s amenable topre',.ntation in a: hantdibook, But the concept involved-thiat of delijer-Sately invest iqatiig natural phenanena and reportiing the results in what--v'er form is requi red for niaximu'n utility to the engineer-can be applied

with equal validity in any scientific discipline.

6.2.4 General Observations, It must be remembered that all identi-

fied RXD Events have two common characteristics: First, as a consequenceof research in sccience or In technology, some new ktiowledge was madeavailable. Second, this new knowledge was employed by an applicationsengineer in the design or development of a weapon systein or piece ofmi I itary equipment.

None of the findings or observations presented in this reportshould be interpreted as claiming that undirected research in scienceduring the past 20 years has not produced a great deal of useful Infor-mation. Rather, the findings suggest that the transfer of undirectedbaslc research to technology for weapon-system development can require20 years or more. This-transfer of the scientific base may be Influencedby early recognition of a need or technological opportunity, by effectivecoupling between the scientific and technical communities, and othervariables -that require further definition and analysis.,'

"The data from this study indicate, however, that there does exist ashorter-time spahnbetween scientific discovery in directed basic.researchand its practicalJUtillzation than between undirected basic research andits: practical utilizatlon; this should be expected. The median time be-tween the occurrence of scientific- events. Identified in the directedbasic research category and the Incorporation of the resulting new knowl-edge into the design of a weapon system was nine years. Thus, it isevident that a motivation toward system application Influences the timefor-transfer of basic research events Into system development. Thisobservation in no way derogates the continuing requi rement for expandingthe frontier in the scientific disciplines.

It is clear that, on a time scale of 50.years or more, undirected:scientific research has been of.immense value. Without basic physicalscience, we could scarcely have developed the modern technologies ofnuclear energy or communications or the current electrical and chemicalindustries. None of the Identified scientific Events would have beenpossible without the use of one or more of the great systematic theories-classical mechanics, thermodynamics, electricity and magnetism, rela-tivity and quantum mechanics. These theories also played an importantrobe in many of the technological Events. For example, If one were tocount how many times Newton's laws, MaxwellIs equations or Ohms lawwere used in the systems studied, the total would far outnumber therecent RXD Events identified.

In less than 5 percent of all identified RXD Events, scientists orengineers were called In to "make someone else's sol'ution work." In

99

R

"85 percent of the cases, a fundamental problem was presented to thecreativ6 individual or group and a new and successful s6lution was con-ceived. The evidence Is so overwhelming that no question 0hould remainregarding the role of the scientist or research engineer vis-)-vls theaplications engineer; neither can be subservient to, or unaware of, theother. Each performs best when free to cope with the fundamental prob-lem. Any environment that enhances coordination would encour, ge thehigher payoff of research. These general observations, by suggestingspecific fruitful actlvlles, impllcitly recognize this fact.

6.3 Research in TechnoloQg

04; 6.3.1 Subclasses: General and Restricted: Of the identified RXDEvents, 91 percent are reasonably classified as research In technology.The characteristic that distinguishes that kind of effort from the equal-ly Important but somewhat more pedestrian work of applications engineer-Ing is the relative amount of technological risk involved. Before anRXD Event could be so classified under Project HINDSIGHT, there had to"be evidence that the performers,. though confident in their prospects,realized that success was -not a reasonable certainty. More specifically,experiments to establish feasibili'ty must have been designed and actuallyperformed. "

Again as in thecase of directed scientific research, two signif-icant subclasses'of hlghlyproductive Utilized research in technologywere observed, both Involving beforewthe-fact recognition of a-techno-logical problem. They differ.peri.marily in the manner in which theproblem area is identified. For~purposes of classification, they aredescribed in this discussion- as "general" and "restricted" technological

* ;research.

General technological re earoh: :,In the first subclass, theperformer's attention was focused on a broad spectrum of functionallysimilar general-purpose technological building blocks. Outstandingexamples included work supported by the DoD laboratories that led to a"multiplicity of transistors with selective, highly. predictable charac-teristics and economlcal:techniques for their manufacture and production;small compatible passive electronic devices; ferrous and nonferrous

Ss'tructural materials and fabr~icatilon techniques; liquid and solid rocketpropellants; and explosives.

'Restricted teohnotogioaat research: The second subclasscovered the development of devices or techniques with fairly restrictedapplications.. I.t included such items as the investigation of missile-

* launcher design, the development of engine and missile-motor components,and specific missile-guidance components of inertial quality.

In terms simply of the number of RXD Events identified in each sub-class, 30 percent of all the 'technological-research Events studied weregeneral, and 70 percent were restricted. (The matter of the subclasses'relative value is discussed in section 6.3.4.)

'4' 100

j4"

6.3.- Mandgement As.pects: Differentiating between the subclass.,es."truni a management viewpoint ,we find in the "general" subclass a lowrisk that research results will not 'be used as long as the product int

U. any way represents an advance in technology. But, without very close

coordination or centralizeda managenment, there is a high risk of duplica-.tive research efforts. In the "restricted" subclass, on the other hand,there is a comnparatively low risk of duplicative efforts and-a high riskof nonutilization unless the applications engineer is aware of the re-

•; suits and they very closely satisfy his requirements.

The ideal, efficient environment is one in which performers ofrestricted technological research and applications engineers are gee-

, graphically or organizatkonally associated. Thus, general research can

be performed most efficiently under centralized management, and restric-

ted research, under decentralized management.

The value of-decentralized management for the restricted subclass-is confirmed historically' by observing the absolute magnitude of theutilized results oT technological research in the course of engineering•

development. From 20 to 30 percknt (.varying among the weapon systems

studied) of the requisite new information in restricted technological

research that was used had been generated during engineering development.

6.3.3 Kinds of Payoff-: -The examples of ,payoff from general tech-nological research form a readi.ly disbinguishable pattern. The mostobvious characteristic is that the technology tends to have peculiarlyttmilitary flavor-or, at least, at the time the Event occurred,,-interestin the technology was found primarily within the Military Services. Butmost of the utilized technological research in guidance, radar and

6 rocket propulsion was directed and performed by individuals who wereconcerned more with the technology than with the development of any

T specific end-item system.

In each case, however#,,the performers were-sufficiently aware of-the technology's deficiencies in terms of foreseeable requirements offuture- systems. -This awareness extended to the knowledge of-quitespecific desi'gn criteria that were ,desired. For example, those groupspushing microwave power tubes for radar knew more than-the simple factthat devices with higher output would be wanted. They had a quantti-tative notion of just how much higher the output would have to be.

In retrospect, it is clear that this quantitative information must

have enhanced the realized research efficiency, If only uy automatically

ruling out any technological approaches (even if otherwise successful)

'I whose potential 'was inadequate to satisfy the criteria. Examples* (other titan the amplitron microwave power tube already mentioned) are

y.4 the hydrogen thyratron, inertial-quality guidance components, liquid andV solid propellants, rocket-motor cases and magnetic sensors.

101

It

ýR ýZ ~~~~A sli g h tlIy d iff t!rent, i IIu s traLot io o f ut ilIized t echnolog i ralI res'narchis 5f1OUnd if) 0hC deVelopillein t Of tanidlum capacitor% and high-Lore-pe rmle ab ilIity inductorý-.* It wasý obvious to key personnel that the nmall11-ness, of the transistor could not be fully exploited if lite size of the0ircuitL cliass is were- di ctated by the use of couiparm ive ly large conden-sers, coils and trans fo~rmers . Bec ause these people and their. close

Vassociates were vitil I? intere~sted in reducing ý-he size of electronic* equipment, te.0hnological research to that end was successfully pursued.

I t may be.that other research eng ineers , .more remote, frain applica-tions engineering, also recognized the problem a nd achieved technological,success along .other approaiches. If so, no evidence of those results'utilization was found in the HINDSIGHT study. there may have been tech-h ological succes s, but'the result s were not "sold," and the researchengineer' s remoteness f ro m the applicat ions engineer cannot be disco~unt~edas the reason for it.

In oveT 93 percent of the identif led Events in the technologicalresearch category, available informato supot th ocuin tacontinuing interaction between research and applications engineers I'sesseiptia~l.to a productive technologicaV research program.'

Incidentally, .such exmpr les as the microw ve power tube, the hydro-gen thyratron dnid.the tanit~alum capacitor point out a very valuable con-,tribution by the .tec IhnicalI. personnelI of the DoD' l aboratori es'. *In eachcase, the greater part, of the tdichno~log ica 1 reseairch by .f ar was accem-plished in private hi iustry, generally under contract with one of the'Military Departments. The work was distributed widely In time andpl~ace. The only names that appear with any regularity as having beeninvolved technical~ly are those ofa few Individuals from thie ArmyElectronics Labora~tories'.at Fort Monmouth, New Jearsey,' the*'Navy Elec-tronics.L~aboratory. at San Diego, California,'and the.'Air Force'sAvionics.Diint W'lght-Patterson Air Force.Base, Ohio.Iticlear that these few people prov'ided the continuity and'effective

*-guidance that resulted in the high payoff of technibloglical research.

6.3.4 RelatIvii Value of'Subclasses*: It has been noted that' 30percent of the technological Events fell into the general subclaiss.:

According to the analytical. technique used in: Projezct HIINDSIGHT, itis assu~med -that each Event deserves equal weight, or value. This'assumption is, in fact, valid only In the context of :a s Ingle weaponsystem'. 'In that case, because greatly improved weapon sys'tems appear t

61" be the synergistic consequence of ma~ny (usually on the order of 100 to200) separately identifiable Events, each contributing a negligible '

degree of. improvement, the ave-raging process-is reasonable.

When the consequences are manifested In a number of Weapon syStems,which is the tendency in the case of'general Events, the true value ofthat Event must be greater than that of one contributIng only to one

102

-sys Iteifl, Theoretically, of .cotir,,v, ifr every weapon sy'-Aemi anil equl pillentin thc Defen-e arsenil wvere analyied and the contrihutltig Event% idenlti-fied, it would be possible to assign to each Event a qudntitati.ve measureof value.

4 During the Project HINDSIGHT study, 20 systems were sampled with"varying degrees of 'thoroughness. On that basis, such quantitative meas-ures cannot be made. The sample size Is adequate, however, to measurethe relative efficacy of general as opposed to restricted technological"research. Over 12 percent of the identified Events in general techno-

logical research had an impact on more than one of the systems studied."Where this situation obtained, it was generally found that results were

A utilized in three or more systems. With respect to more restrictedtechnological research, the use of results in multiple systems was rec-

p ognized In 8 percent of the Events, a' i then infrequently in more than

two systems.

6.3.5 Planning for High Payoff; ihe many observations on charac-

teristics of utilized technological research arising from the HINDSIGHTsystems study could be used to provide guidance for senior researchmanagement. Where anticipated results-of the research are consideredbroadly applicable and the estimated cost of the task is comparatively

low, full authority to plan and implement the research program should bevested in the scientific and engineering community interested in the

•;4 pertinent technological area. Conversely, where the-apparent applica-ýV bility is more restricted and the task's estimated cost is relatively

i higi-, the interests of the system-oriented engineering community shouldbe dominant.

SThe chief possible exception, of course, Is a situation in whichb the Initial cost Of acquiring a facility for technologicalI research ofgeneral application is very high. Management must then judge whetherthe application cost Is properly attributable to a few tasks~or whetherit should properly be allocated to a great number of research effortsplanned for the future. . =

Again, with this general exception, it appears that there is greaterassurance of a hiigh payoff from the more expensive research tasks If thesystems-engineering community has some responsibility for research plan-

ning. It was observed in this study that some 20 to 30 percent of re-quirements for new technolpgical knowledge-in the case of MINUTEMAN II,

up to 75 percent-are established while a weapon system Is being devel-

oped, which demonstrates that, at present, the systems-engineeringw community cannot forecast all its needs.

An apparent solution is to have greater recourse to the current R&D

it category of advanced development. Useful prototype weapon systems-should be designed, developed and built for the express purpose of

giving focus and spur to the growth of technological knowledge. Thisconcept is not new. Whether fortuitously or by, intention, It has been

used to great advantage, and its impact Is seen through "hindsight."

103

hi Io

Thus, in retruspect the- tremendous value of the NAVAHO missi ledevelopmei)t is recognized. There were a number of reasouis that theNAVAHO. never became operational, paramount amung which wat thie appear-a an ce of the jintermediate- range and intercoritinental ballist Ic misi les.

, Work on the NAVAHO continued throughout essentially the whole RDT&Ecycle., however, and did provide a quantitative focus for all the tech-nologies required to support its deve lopmenit. It is also noteworthy"that, at the start of their development, both the ballistic missile andthe NAVAHO were considered high technical risks,

Studies of the technological basis for modern guided missiles (suchas-MINUTEMAN and POLARIS), aircraft.and nuclear Submarines almost in-"variably'find their way back to or beyond the NAVAHO program. Inertialnavigation and gui.dance systems,.stellar navigation systems, flightbornediqital computers, liquid rocket engi~nes-all have been traced back towork identified as originally done for NAVAHO.

In a ssimilar vein- though less dramaatic and on a lower cost scale-a study of the-history of the LANCE missile system establishes the value jof the Mlssies'"A" and "B" programs In the 1950s. Agalh; requirements.to advance the sta~te of the art established'by operational specifica-tions for these developmental1mlssiles spurred the growth of relevanttechnologies,".

Amore specific examplea: is the contribution of the Army'sn DARTantLtank missile, which was never produced. To be economically feasible,the DART's roll-reference unit had to be provided for a fraction of theC•ost of available gyroscopes. As a result, part of the technologicalwork on Inertial guidance components was turned away from the quest forever-increasing precision and focused on the need for marked cost reduc-tion. In the absence of such motivation, it is unlikely that the quitereliable, r'easonably precise, very Inexpensive gyro used In the BULLPUPwould have been readijy available for it and for other missiles as well.

These examples and numerous others that have been identifiedthrough Project HINDSIGHT studies Indicate the wisdom of undertakingsystem-development projects in a conscious attempt to focus exploratory"development on programs of research in technology. Advanced systemswould then be described:rn terms of operational characteristics andcategorized as advanced developments. Maximum freedom should be granted fthe project director to seek advanced technical solutions, and at theoutset there should be only limited definition of a production objective-preferably, none•

.- In principle, the planning inputs required for a high payoff of thegeneral and restricted subclasses of technological research are quitedifferent. It is possible to Identify real-world problems for theformer by examining classes of weapon systems-submarlnes, aircraft,missiles, armored vehicles-and recognizing that the growth of specificareas of technology is essential to the marked Improvement of one or.more of those classes.

104

FL)r I eX.mlp 10, it is obvfous that*'a subm.ative'%, depth• lerfoJrm~arfL it,/

lI Indted by the -ifren~gth-to-weight ratilos -of available prr•;•ure-hlull""•-°-materials and by fabr'ication techniqlues. For ai rcraft•, haievmn

Sth

of greater eff e' ct ie oplerating speed depends upon the availabilit y ofthermally less sensitive skin materials, as well as better understandingof aerodynamics technology. To gain greater precision and longer range,artillery weapons require.more efficient propellants, reduced inanufac-turing tolerances in projectile characteristics, less tube erosion, .andso on.

Because such problems are relatively easy toi identify and researchresults may be applied to many systems. the inherent probability that

F the knowledge will be used is high. .Further, duplIcation of research* can be safely minimized if planning responsibility for a technological

"area 'is centralized within a management that is thoroughly aware -of- general system problems,

"When the technology has a restricted applicatlon; real problems are"not so easily identified, The risk that random research wi-ll yieldresults that are inadequate to solve any specific problem exceeds therisk of duplicating effort, and it is suspected that the work ls ire-

-•'.. quently of less i nterest to- the performers of general •€'ientifican

engineering research. In order to gain a high payoff of restrictedtechnological research, planning responsibili-ty must be decentralizedand assigned to the specifically interested applications-oriented systemdesigners.

I -

IK!

105'1

A.,

b .

"Preceding Page Blank

I. REQUISITE LEVEL OF INVESTMENT IN RESEARCH

7.1 Stratetgyand Fundijn

7.1.1 'Research Strategies: Every organization, whether it be asmall company., a large corporation, a nation, or-a national agency suchas the Departnent of Defense, has a research strategy. It may be ex-plicit and well documented, or it may be determinable only through In-

* duction based onobservation of,;behavioral patterns. Nevertheless, thestrategy exists and, more than any other singI-e factor, defines the

. requisite level cf investment in research for that6organization.

At One extreme, research strategy can call for assuming a"I 'para.ki tic" posture. In that case, the organization allows others tosponsor and conduct research; then it adopts (or adapts) certain of thefindings in accordance with its needs and purposes. The cost of thisattitude is measured i~n terms of patent fees and licenses--and, perhaps,in loss of co~ipetitive:position. Such a research strategy is typical ofmany small:compan6ies and also of newly developing nations.. As reportedby A. H. Rubenstein, 28 it 'typifies most of the home-entertainment elec-tronics industry in the Chicago area.

Further along in the spectrum Is the strategy of "reaction," in

. which resources are allocated for research only to solve Immediate prob-lems. Great scientific discoveries are unlikely. e.ither-to be made or,if encountered serendipitously, tobe exploited. Radical technologicalinnovations are neither sought nor expected. This conservative strategyseems to be acceptable to most consumer-products industry in the United

. States as well as in some of the'smaller European nations. Expenditurescan be modest'and research efforts, highly selective. Appropriationsneed onl.y be made.after a problem has been identified. There may beadditional expenses for licenses or patents, but the probabiliity is some-what less than in the case of the fully parasitic posture...

The reaction strategy, if tied to a well-planned, long-range engi-

neering development program, does not necessarily incur any great riskIof loss in competitive position, national or corporate. If observed toorigorously, however (as discussed later), it mitigates against the Intro-duction of markedly new technologies and so leads to rapid increases inthe cost of advanced engineering development.

The antithesis of the parasitic posture' is an aggressive, forward-*. looking strategy. Clearly the most costly in terms of immediate cash

layout, this strategy prescribes that sclentific and technological

2 8A.H. Rubenstein and D.E. Brewer, Research and Development in the* Chicajo Area Flec-cronics Industry (Evanston, Illinois: Northwestern

University, 1962).

107

Vi •.

•-;.

_____________•____

o.portunities he .n9grre.,sively ,ought and, when founid, exploihted a. .on-!AIdvred necessary or desirable. In prtinc iplc, this strategy ivuids the

S- risk of losing competitive position atid minimizes the need for coustlylicensing airringemets. Its great disadvantage is that many finodi ng,will never serve a useful purpose, because they lead either to undeiirredcapabilities or duplicative teulniques, of which only theit ore attractivealternative will be useful. In practice, the forward-looking strategyin its ultimate form will probably never be adopted, becausý unidoubtedlyopportunities for research will always exceed ability to stipport therequired programs.

7.1.2 Strategy for the Department of Defense: The fundamentalconcept that the DoD should support some research in both science andtechnology Is not currently In question-. It is generally recognized"thata potential enemy who is technologically advanced (or advancing) repre--sents an increasingly severe threat, and thatconsequently this nation'smilitary capability must be continuously upgraded merely to maintain.parity.

The rate of the Dol's Investment In research, however, has beenquestioned. Competent authority suggests that :"The high l.evel of supportof basic work is producing scientific and technical Information at such-ahigh rate that it cannot be'effectIvely digested, Interpreted, dissem-inated, or- put to useful purpose." Translated.Into terms of the three

strategies discussed here, It Is agreed that the DoD must adopt a "strategy of-.at least reaction, -but it is felt that -moves toward an ag-gressive, forward-looking strategy will require a more pragmatic ratio-nale than has previously been afforded. This Is inferred:from the fact."that critics are demanding greater proof of value In relation to thecurrent-level of expenditures.

In reality, the problem faced-by Defense RD management is twofold:First,. define-a research strategy that is consistent wlth the strategyJemployed in- relation to all R&D; and, second, demonstrate that the re-!search strategy defined is a sound one. The requisite level of invest-

-tment In Defense research, then, will be the minimum needed to support-4 that strategy. -

7.1.3 Relevance of HINDSIGHT Fihndings to DoD Research Strategy

and Funding: Project HINDSIGHT was neither intended to define 4 research•" : strategy for the Department of Defense .nor designed- to defend whatever

-. strategy has been used. Clearly, however, many of this study's findings. are relevant to those matters. Further, the HINDSIGHT data permit a

limited evaluation of balance among the several areas of technology in,the DoD's research program.

-,The purpose of this discussion is to highlight those findings thatbear. on matters of research strategy and funding, interpret ng the data

where posslble, or, where the data alone are inadequate to support a

108

:-- .A

SC.ucI u., (i ll . U I-CuL l [it)q oi L tle i r proball I e • rni i Icance. The I uq iC i .s a

It Ctetljyidhr the Iet Urn on investmlrenrt ru',ulltiriq from Ihes l'ti"It Jy thalt haIs been21 Csl)ousOLI by the• De(partlilluit of De Ifllns "ifire 15)1ý;

(2) Con-paIe the DuD's resuarch strategy with an averageds t rateqy of tlhe i otens ie Ily technIogluical U. S. industries; and

(3) Examim: the Ii ke ly consequences of major shi fts in DuDresearch strategy or funding.

7.2 Measuring. Return on Investment

During the period 1945-1963, the Department of Defense spent approx-imately $10billion on scientific and technical investigations. (Thisdoes not include a considerably larger sum that was invested in the en-gineering development of weapon systems- and other military eq uipment..)The strategy adopted for the support of this research program can beinferred from several observations in Project HINDSIGHT.

The most significant relevant finding of this study is the determi-

nation that 67 percent of the RXD Events affecting the systems examinedoccurred prior to the designing of the equipment through which the Eventwas identified. That is, at most, one-third of the work could have beenundertaken in response to specific problems uncovered while the system"was being designed or developed.

This plainly suggests that a considerable portion of the researchmoney-was in fact invested under a forward-looking strategy. Supportingevidence is found in the definition of the R&D budget's six-elements.Although this budget structure was not formally introduced until towardthe end of the period studied, It served more to dignify existing prac-tice than to impose a different research strategy.

Moreover, according to the historical funding pattern, from 20 to25 percent of the R&D budget has been allowed for scientific and techni-cal investigations. Any measure of the return to the nation from theDepartment of Defense's 1945-1963 investment of the $10 billion In scien-tific investigations inherently includes the measure of a forward-lookingresearch strategy.

A crude but useful way to measure return on Investment is by estab-lishing the cost differential between a modern weapon system and anoperationally equivalent array of Its predecessors. This differentialis then compared with the cost of the science and technology that enabledthe advanced system. The posit is that the point of marginal return hasbeen reached when the cost of new research exceeds the savings gained byreplacing the earlier systems.

109

.a.

~~~~~ tA I3 t'.'b ~ .f

'tit; t' 13 1 t31 . vi h13, ~ ~* ~

1 11" .3 o~l~ l fjl331 1 ljt fm t3'l ll ll) M3~ i'3 3V [h ''.3.~ld

of3~' 1133'3.1 '3l 3 i d' It il~li I'lt of3 I It 3 i t. .3Il1'3 I,.I 13 [1 v31 3' 3 ''

11VOI~., I !M it lltt. I I t3v Y llklv

ot'. pfrOv i dtie J CO333p I t, t v I y now~' ca a 11 I I y t33 Il'It it,) 333333v (if t ht-1 1 (Il t1'-s'i V II3 33' .N C u dId 33d1tO.h , 0 1'of 33 M IMI * 3 it A % 3332.3111 fit 31 1 -. to- Lotillpli 3'ity tltu333,

Af: ~~~~~~be r olf t IhI 90nmlul 'or1 1 7203m33l Itld i t11331 or I I V.1y a I I i it 3I 1- (3 .1 V 1 a t it e. ofIAL 9-'wi3th1 lif xt-c 30rt'3cont3 NIRE -IIERCULLS mi- .si 1le dultve The3'~113 gull'. wetrv%o0 1 i fiti L ed witIf 113 wct o 3 2i23 I0 13 Ui tud t ia L it i f Or i31 13. 3U, i 3 r C) b dttI.Ir i 1'

coul1d [lot aff ord equ~ Oivalecy. IFor Ifl3rf)s1Ž of the return-rth03 3Vtivutwl33

i na Iyst's , tie system3s are comp,,ired onlIy on) the has i s of capab ilitIi Cs t hiltbuth posses; exces-sapabl~3Iity 031 the part -of the s-ceisour. is treatt-das a5. bollus.

'.'~,because it is esstwntially im3possiblt' to ideoli fy andisola3te the upported research in sciece1C or teclitioiogy that contri buted1* olely and total ly to any given weapon SystLem, a co!, t-vlIutý -analIys is ofresearch in relation to-a single system cannot. be performed. It iS:possible, though, to estimaLe the total DoD investmtent in research3 during

Jf the time the k%,~wiedge was gained that enabled the slistem's developmentaind, cnen, to cons ider any cos t advantage of tLhe new wpapo31 system in

,tI7ter3'1s of a percentage of the total investmenlt.

The Task I stildies of Project HINDSIGHT hive demonstrated that tilenew knowledge utilized in modern weapon systems was gained predominantlyduring the years 1945 through 1963. DoD cxpcnd~itures for research inscience and technology in that period have been estimated as betwe'2n-$7.5 and $10 billion; the spread is caused by'chainges In accounting pro-

AW cedures that are believed to have obscured some of the expenditures.Hereafter, in comparing the savings accruin~g from each new type of sys-ten, the 1igher fig ure will be used.

Thra difficulty arises, a-s a result of various designcompromises, that have been made i"n a succession of' weapon systems, Thisforces a certain amount of normalization of system designs before jcomparison cat) be made. The fact that the requirement for a considerableamou3nt of new technology continues throughout the engineering-develop-mtent of an advanced weapon system indicates O:at the system'engineer has

110

II- ~ ~ lif.-I-

I N tt I tIt I -

II .ltstly V f t L tmit t It If Ilc it

Il t, - I i~iilt t i A i ti For1ce lit rouct idp tinC10*trisot lia 11)lcy Ir"rI ft I tt uk

v .l tt IllV 1 tt k'l(t the C- I 1 111MM1 Ii;r f w I Ide to Wil i .1 opr1 Vn int110rV Irjc ,l byS I G hl III", ; invest. 11 i Iae th i I IIeIret lc InI t lt- tII:-.

otiv I .o I nd tt ý yi Ir the I [, dev C:oa ie I. of II]Ithes. e i twi o I I C rIf artd I st ah Iach Ie~ wi Iic thI tecnolqi aviI In I99 Thog rt fi t I th1e1CLI('1

A 1od 0JI I(yanO tipyWde CI 30)I1L~ L ad usel"ll I f h t I cchnv 1Ioo i1 1a knowedg aIu rdurin 111 1 910-yW il Cmat liio h basi pti-rcraft cp de ig. Moifticati n I i y.rI roi i poe I'CJ1"11r. IIgkrIjIIi fI c t I F1 tain on uprdig suc asV. thnd

compt)letel redesigning thle dvlI1101.o h w aircraft. Dept hs iiationstabihe dprhdutIfivityorint caatei~ of thle C-141 wacnresdou ld pcc not handte raeng

acul! poene Ihtialeo technollogy could able exitd An a95 - h o nsh equncCli 's I. ,6 cota pe datg en hti e ee prting

pEylodI(adndiffraenta in30 exces of of0 percentloandalanoeffetie rang edincre tse of tesatn order of tiemanitude wrci civd n diin hhour taf deied ala ve ie itneignd this diniail byreduce';L moverall operating cont Is A very caLio onsegrva iveetmt n f uha theextent tor w i chetfuselag in l sc enstifcendtch nol(] o gi ca eerhicrcomp~r inLl rdsigin the C-4 a rcraft ta h -b Deonithe rhoe laive patiiia, otheperlodut-ivi aah ity of thle C-13 was incrantshuged C-ou 430 vcntndte r ane

';ie) a. entttrj i rely npera de igni waS s. s unetaket asup or t h C-4 , fiiql -r eFul*l~ iotentiai of tecnoog (pr c oultdi ILie expoticoted Atsr ton-si quIe (nce, ait

payoa diferntal n xces f 100pecen ad n efetiv rng

Inraeo h aeodro ,an uewr c vd nadto, h

CS4' .xcma pe datg*menLa:i ae ee prtnhorUodlvra ala vragie itne ndtii un

VK

** (-9

I +'K

� *.flV.r.flf. 4:. 444.4 444 � *dI..441

4h441

4) IA

Ag

p4

444� 44 44 �

*

Reproduced From

Best Available CopyfrI-?

I

.2 11 .1 I o, hn,, I , ,I- I . I ssl. it ,1' . 111. 1 2 l. .5 . I! II I I

SI Ns I I2 s 11 t ' I I ( W)1 . I I III' .1J 1 ho I - 0. 1!1(1 r. 111 1 1 .1 1,i

t ii I 1 o o51 2 1 5 1 I IV I i d e .. I i Iw t ( 1 1. 1 1 t 1. In ~i , I -I I vi t i ..i I l l~r,s s

u . 1, hIt it Iis -.IJUi~ lilt al, .12 tIltia diitjs I . 1J~12' fj 1511,4 I, h 111,1f i )'5'.2!2u1 5

I tli i Is i I d IIt l s I I t c I~u I Iiv a i t I I II Ii, Ai 'sIl I w .Isi 15 ij I, I

to bu 11 iif .. 1ik Ill' vilI'sIch ' o t eili6iia i sI qyI If I i , I I iv siq%1 1 1kIl i b lt-I o, II- tl lI [l IlIIIO ,i1 2ty' hye opt iwII/iII(q thU!((" ql 0 tln ýi o l II) - i- o s o i t I e llo I ly [ttIliIIii( t

01 VNk ' Cs cOil CIUdktfs th at , in 1 Lit!e lbt';llue ut tI. t' C- 14t 1 t tIII total mii 'lss

L'JsU ILI bes 1.101,0 L'Xpeilnsv IVl)5LsI-U'st- the C-13t$ (or i Ls IC chnolo I oCi i aIei vIii-I01 I nt)WOol IIA IVC to per form both t ht' mlission forl whiCh it. i opt i mi.t5and tiald for which the.C-)41 is designed.

AA:.~ ' t./;P.I;-.IP; i-Iinri Another weapon 'systLem studieduinder Project I-INDSIGHF vias the AN!SPS-48 radar. This surveil lIance andt a rsjetIacqui si t ion radar , now beinly used i n the f leet, i s general ly anoperat ional sceso to, Elhe SP surveill Iance radar of World War I IThe re- have been other intervening radars, but. Iin "errns of technology allIwere subsequent to the SP. It may be assumed, however., that; tney u'iedpart of the i.-!w technology purchased with that same SIO bill Iion, so theyneed not be considered in this analysis.,

At least 40..SPs Would be required to afford the arne tJegree ofeffectiveness in radar surveillance as that. obtainable from a singleAN/SPS-I,8. Because of significant differences in the two radars' maxi-mum range capabilities,.39 of the 110 SPs would have to be carried onseparate ships, optimally distLrib~uted over a large area, to normnalizespace power density and data rates over the volumetric coverage of theAN/SPS-48. Let: us assume that t his could be done and that adequatecommunications could be arranged to maintain the distribution and enablethe effective transfcr of information.

The acquisition cost of the 40 SP radars would be from $6.5 millionto SIO million more than the singlc AN/SPS-it8, depending upon how muchthe unit cost might be redtliced in. view of the greater SP production.More. important, though, 39 additional ships would be required, each witha total operating crew of some 200 and costing in excess of, $15 [. 1 ion .Assum ing cos ts of $200, 000 per SP aind $ 1.5 mil InIj n pe r AN/SPS-1i8, thledirect capital-expenditure cost cif acccxnplishingj with 1945 technnloqy

113

BEST AVAILABLE COPY

.:ow

XV

ho I o-. n t . 111 t .. ~ , 1 1- i fII ii 111 1tt D I !;

I~~~~ ~ ~ ~ tO t .i t1

1 o~ a 1 I -e hI vo I I i t- i s r1w edp i mnil It~iill It-w u 'd t,1 - wr iv 111 11 #-

The capabhilit of filte ol 'ter radar (Imihl ittel havtI leli~tit Ie pi i It-dI'~iIIW -.01 i tltnl inPt-et p ient II hIS Vdldr example i'w.' to vIII1~t v ii '.n11l

OW 111I I Id l. l t Ld titl O theV lay~ff Ofr rtLsva rch io # ient ýWt! t, ItPhe !,irIier e' di'pic if 1 he ai-rait, on the other-1 itatd, i ti 111 i,crete. rexil-worid ciist.

~''e;': The. retiurn on i uVet'fiont i n rv~ua I-can he timi larly demonstrated for each ot thw systlem% s tidied. To thy(parLiticlirly for expendable -itemts such is miss i les, mines, torpedov. oi

other muni t ions) woulId involWe the use of highly classi 17led infurataitaton expendi turin-rate planning. I n somie ca-ses (e q . , nuc lear wat-heod-,even reaieefcleesratios are classified. An indicalton ;Jrealized returns fromt new 5cience and technology in sa.me ctlhe',yiehowever, is Suggested by the ratios5 shown in Table:XXVIII.

Table XXVIII. TYPICAL INCREASES IN RATIOS OF OPERATIONAL--EFFECTIVENESS IMPROVEMEN1 ENABLED BY

TECHNOLOGY

- Estimated factor ofWeapon System -operational improvvement-

Current -. Peeesrin a common role

M4-102 1 O5unv Howi tzer MZA! I 1"atiim 1w i fI .

M4k 46 Mod 0 Torpedo M4k 14 T'loP(dn32 fM4k 56 Mine M4k 101 Mini,0.Bt1111 PUP ASM Free-Fal i i.omit 4.11

BEST AVAILABLE COPY

-4i

itit.

:1 " i f 111 1 i .11 1 it-, 411.1' huv -t a It Id e I I ti itiit

Ien .k I 11o Ie L velni ' lI i uq~ty

to L' I eIIlI 1w wt do l ., s I I d II i oi cIaI ilive dI sl uml I ha if nus ,Il t - 1.111, bI'a,1.ii r.1 -o k~l the -DoD f re'l' progra L vly I h Vt I t1 11be , f d cll* r'

L Ie Iu av I lt Iar reeac 11111 ,~n andt - %I t he re 11 atU~ I re Ipo tj i v.

ri~dt-at 11 un(!;ad miltl conipe t~ in .q sci ieiies and tech 1101og i eS Th is studiyJiki.. not ex tend toadirec t tes t of thfe wkdoni reflected in total dollabrlevel I, l.P~ h owever, eto sIimate the quali1ty of j udginer t.

exeic;ed ill apportionmonts. Onc Mi ght thecn pe~ul~ateý that, becau~se thesaepeople were invol ved if) both sfit~s of juldgments , the wisdomi di Splayed

wa! ab0LJ as 9ood iin one as i t vias in thfe other. Also, an i nrli rect Lestof tile r~snblyof thfe total dollar level is presented in section*7.4.

in Table XXIX, many of the sciences and techfnologie5sShowil in TableVI II a re ICstLed i n relIa t ive order of funding during -FY 1966., Someu i tLents

wt:re om i t ted because. the systems i n wh Ich those t echnolog ies would bei..1poctalt. Wei v net. repretent~ed by the typc!ý stukdied in Froj.ect .HINDSIGH4T.

Td'ult% XXIX. RELATIVE FUNDING or SCIENCES AND TECHINOLOGIES,' FY 1966

1. Missile" ~ 8. Physics2.- 1'aiga'iqotio nd 9. Mechanical and civil

~JjrW1'Ii Cd: i~l3 . engi neeri tig3. dna. ~10. Materials

t .tr':rjc it 11.- MathemiiticsII !.11c. aild Fli:I'ls 12. Chemi stry

La13. Atmosphei Ic/ :. N. i-; vd vu, 1. Inent 14. Enurgy cuonvuts inn

115

BEST AVAILABLE COPY

.4" 1 1

I'. 1 .Inc 1, 1 1 tl 1s fie I ifat'A.i I'P I 1. 1. d i i. .

.-I im t ot r hsilk.1 i)~ o. - 1) .1i 1i aki cI of 11j1(r.1 lt- A I';ll- tIs ci ined .I ; It I~ L oq i ha Ir Ii Ir in off til it ).t I vi ltNI) I IT R ~I ,i V% I' IO'. 1 i t-1lv Ila I di i

.I I ni~i io Cal" ofi l tit on'.itltr i~ f Ow -V 1~f .. 1 /Ili

Nom t V11io *itialc. tha)t tliist Iar iua et. fof( i int 41100 iinI -- it i'a ofill, sf rn. 111 -196 It Ou'ly ' I i-, v I 1nq ptiij t. INWn Itha i % t- as .iet fled -

efot of lwyavy fund i riq ,tilvrt., have 'bton more accump 1 i -;hment s i i 'cu:r I.,- iniii ea, z't -.c~iotnce and tvchtioloqly UnWdoubt edly i hv re inutt tic %Iome cor re-(LaitIon btv-won levetl of fundinrg and probabi I I ty Of succSsful _aJhieVe

10on :t t-iuno deicret-, tlti:re is a resul tant bias in the HINDSIGHlT fltau.Iin ctoeir'al., flaweJvei, the appro-ach used in this study-the ret rusp'ect ivint razint of the flow tif !;eience and tcchnology, s~ataLing with 0 ti I i led.ra thIer than I abora tory-advert i ed produ cts of research -terves to minti -tulize thle consequelnces Of' tile Wias.

To thle extent. that managers of DoD research dictated the nature ofthe weapon systems to be developed, they had opportunity 'to encouraq

tieuse of their Programs-' results in those systems. lWhere-ver theweapton system was% developed by an independent agency, such as an indus-trial contractor,,the opportunity would be minimum; and that was Life-:ituation with respect to almost, allthe weapon system s studied inProject HINUSIOFIT. Timus, thv ran~k correlat-ion does measure the qufalitvof rmanagerial judgment in at least: the matter of resource allocation,

A% previously noted, a co(efficifnt. of rank correlatioi- of 0.1 -,P(:tl.6he "hirgh. "lInoring the pos,;ibility III logical adjustment to 0.113 a.-Ie',crih bed before, a corre Iat itn of 0.33 i.. enough to demonstrate thaiteiianaujrmen~it lwin. done iin excel ltnt jot) of atloncating its resouircos :ii,

cilitpe't i n scientific disc iptiu Ie ndl arvas of, t(chnll olqy

JCo rrelIat ioii ~.gn ificatce tost% I

BEST AVAILABLE COPY

I Iq 1 .I..*,

I ~ ~ ~ ! w. y ' icl l I ii lI' .in Il ai I ti, o I:;V,L eIr eit .

IC Iitsl~~jI alb,'.111t1' V? i'Cfle LI ii w at 1(11' It I1dU r 1 ~~'e*rIII.ii a' 1 IiLiq Ij ew Id' have !,CICL~ f it-i 111,Wi D, ) LI'It'l I wl'. t r

h I L lii ' 1 1i t 0t'a i l . I I I'tt 1 ' 1d. I I'.~il~ IL Ii pu I ;w I dte 111.1 I yij fit I rcli ." o I I Id 1 0II 1 i I t. M". i l I v I ' h i, ti I t-.1 ofe I silp - iIi (ill i 1! it f i t'ro ii111 tl']

v tr Idi 'k~i ul i I (I ii I C I v.; tci -Iii31 Cl, it- ,I If P lI"I li I I i C h IJI tiL LI I-I d joi t I a I andt L'C I' .f I it'( -y r~il r% i (n I rIiw riv -.t- i' r-. .1 1 .11

I`0n0( qi. raI I Y (oi tlln i ve Iha o r epIy nv 1 ved iiIvhn qIcwalact~~~~~~ vIi cs mu-; I tti in L Iqris le -A rcsvrc an deeopet p oniii

1111 JiiiArke 111 .p ILe hod t-a o1 Ion v I ou In, h S I i tLv Inc Th Dea Irt ivltent I- DL eI'

-CI t.Ii n ta i -,ul !nalogous i ituationC o~nl wheni 11 inCerlatioa peNC t cc i I tin.-

I nu iczlnqor. xWhipni heris noa ai.ý dedgaep v theat to nat CIonal sC_1

:tmy be poss ibll. to curtailI or ptostpocn the development or new weapon

Ieliw'. tint il, as technology growis, opportunitLies for even better we'apon's-iri';e. i IIF the DoD researchli proqrn is. malintaine'd at a healtLhy levelilur big I hose re.latively peaceful t~imies, the balance between re-tearch anddevelopment should shift away from that adopLed by industry. In brief. .

ýj*

Du19 tie l( pa', I few ducades the nat Iional se~cur ity has contkjitiow~l'Iyi iI it.-p.i rdy . The allocca Iion (if DoD rv sou recs bet weflin ccart n

!1-i'I1-;Ipl;lIn (aIll II'* i'(alli ng ful ly collilpa rtd witLh that of thiil..; i ndwi' t I-i alI

ICOPY

[..

I,', .• pij . uv l . pl it. I ,,n I , l I ,', i i , Ih. th i -I A

[ ol~' I 1w ' I,1• I•.%I)I Itd~ln. l r l c.+l•h .•uuI *.1iqlhl Iy •~vil 14 ijn'•.n l',I I "

L'\~ ~ ~ ~ ~ ' t At! ~y lv'I '

N1, i ki / inil fbi) '.r; t-By a . h.i. for •.Icul.*nion, 1.ihl ' XXX di# p ! 1 1 -.

,hIo fitr yer I 162 th -ouqh 16/: I

r o the -re ortel )I I ,ipport I onment oII r to Iat R t (cot usl -2);lI 1 het I'er -I aendtt res or al lc,|tions IIr resc-arlilI

e cjxpIoratory devI lopuct- Iunns 3 and 6);t

• . the levels that would have been asSigned if the averageindu.tri,,l funding strateqy had been invoked (columns 4 and 7); and

the percentage differences (columns 5 and 8).

Apparently the Department af Defense has adopted a more conservative* posture than industry with res'pect to research and, until recently- at-,ore aggressive posture concerning exploratory development.

Difficulties in reconciling the iwo sets of definitions may berespou|siblc for greater apparent differences in this table thani actuallyexist-that is, the DoD's definition of research maybe more rigoroesthan industry's. The consequence of that difference would be a mis-"leadingly hijh industrial figure for research strategy (column 4) and aco-respondingly low figure for exploratory development (column 7).

" '., oj &zt. o• &tj nec , u :otaec (Washington,. D.C.: NationalScience Foundation, NSF 66-33, December 1966), p, 9.

.."' (Washington, D.C.: National Science Foundation, NSF 66-28, June1966) !>Z.I. '-'yZi P, Jnvr for l(!acl, an,1$h 4.o , t $ , t I..-'. i , .' •.

F7.. j j7917- (Washington. D.C.: National Science

Foundation, NSF 66-25, 1966), j5.

118

9

BEST AVAILABLE COPY

I AV

II.,

II :i.- A**i I,,l e i [& 1 . 1 .1 1 i vq ':I by ;

I~ I ii ;,;: ( I,)I I I I I1 6 1 kIs

I i I. iitu) I (l d VII M' i ll 1 1. 1 L, Is'm L w, i, n9 11 i % a I y', * r or' 111 v0 1,11hiii,'LsItiI cIiL't1 CIu pI vlilfi '., L t , till it? 1i u vs.. l'o vI- eapusr I sy cl 1,j di I ri, plsin- I

JIs i rs'.I %. i t II I P idlt ris co" t i%,Sill I i Liil fr ll' sp. ip t cm i f ie I1t s i oil(J Is o it ex-,ilsAIi l lder i t oo a i I i pmrrly nai IV Iib I IYcld Lh. Iesr ri.. T01i; i .

'ilt Ii '1s'illd !"' p.11-H y .offis't by, the f;ICi that it is iurfc i lipor'diIsiii,:irtl' I iii -jilt-ivavi l (A01 ials ioll 'hos,1 uf a 'i nyle c0lnpalsy wnd, I ll: rLe

fiii,., ii 'Uu)'li. ol Oimii d oidop t oii e.ven wIore, forwa rd - 1ook i ng re -e cl s rt i

iP,/ tho tihoLii mo's[ I ichno I ciica lyb in tens? ive industry,. Neve i tliiie'. I h.n1;1.ly vs is sui' dkemonlSf Lrt hat Use l oe f DoD expeiidi lples frrsa

* r, ~ I *lisI' d tec~hnllIoqy has sIot been sign ificant!ly out. of] inelC wi iiWhaIdt U.S. IislustILI rial ist ruLeqy would prescribe in o~n analogous viivirfun-

/.5 w i-.ernnic-. of' a Fisrwasrd-Looking Resea rch St ratey

1 it I Iho prece'di ng 'sectIion, it Was nioted th~iL tht~e Depri mrtncit oiol~fs-Q~vi d2.nces a forwaril- look-i fig research strategy and fiur i.hivr (witlii-

si' irgsoiss'uiion) that- luch i ,:trjttrgy can be the iiest efl'sct lye anid.. 0,ctF -f tlie sevvrail possible al teinativeis di'scribec inl r'ct iti /I1..

rioditq', 'iI 'tie hiINDS IGHI stIudly 'uggorst thai . if I si easy irlcua,i'10t ii)''J'.t~s ~y-I sis',por n-1111hMjii or cus I -v'ffcci 4,I~vsuS is isough!I. (liii.11i1s

ita'in' -i ripsil iscolatiori of ri!vr'trcll cost i,ý to adopt anl aqivs-~ lioki ncjriseoiisrs i Iraitvrjy (Xll I5inq for I coidisiit-rolilte

'1 I it, I ly yociI(XI iii vi' rt'',i-,uchi.

BEST AVAILABLE COPY

*C6.1A

4 6c D

ILL

-00 j

CLC

VIV

120

BEST AVAILABL~E COPY

I 1 1. ~ *.tt, iJ i I I v. t I~ . i I I I I I i v I1 I it1 I. I 1 iil 11. 1 i ~ 'I ItpI li~.~I t0 I IT'~~ItiI il Il II! lI- * . ii I . o I.t ,- n I f I Id -. pt- InP I I .c i l I I ,I

I.i iýtI I II li I hv' n I i lld I I III t) i~ 1'111', 1 C i W 111)III1 I II IiI '

--.i1ft* o ' I I inial -tt itrvoiy I v ac t or n r I A,. olpruhci..n nJvIcco. In -,I, i I- . II i dt- I I. od 1 v I n.... I a Ii1S1 ifide- Itte p e e e 'A 1.1 1tL*

10 Iica Ip I .t I ellaItj in I a'iI.a [i $1 Ad i 1 n ot IIlt he aunie ul j tX vrit rfill

Let i nedIa>, thei t I vnI tit, tIIhe su10 .o Y,; to v-ef Iec ii nots o th to

1h 11ed 1- n (M r (IIIV I I i -. Ii (lito s It' y flip I~o tll~ i acd it ori or ef Ic lit vrcIle/cu 1"etc ICt* S0enesPa i~ s. Th sophsIo t t '1.1.1iII.t ionI~.L -of thIe pre -de ceso is o I I ( h

tunica1 o l ICat i o or 1 tlito of lite lumiber u f RXD Evcn' te1oqriibiothe , baequ 10of

to . ntuai veIhoIly, th ! at hemalc ti ,calw formultion expressediii in the-plotdpearI'i rea Isoni IahtI. Experf . ienSceC.W tlS Cu% tha gee rly ffcIv ane st an t eqipmefbec pomescss mo Mre opexo spisticated a g0.1sreater amun 11 o ff effor i elý,/(StthaIv :t experlies c may b.Te rephra sed thc impr IovfLiepeentefscor *isprotina ;o some nnlin ear pound rtilrctionf thmde ompl.xt Thr oph I-ticationI'JI .iOfof t he new sste and to veL the levelofcmplxityortopthistiation0 of

Forliyey the firs-orerprxmations perniiatti xrsedb the tvihble daotaitappeard reasonable. t. assumene ato]aIsu thagenerellationshi bnetuieentnuber -ofe RcomElent or spimprovetent factor.e*ther ramounabiltyfoft tisaesumpireon eac supporedbyte facret that thepdotameit. Aalstrightlinetha oerxaeraenge ofy depatrepthratseds ahmrvmn factor of 10ropteovrticnlaxi and aom fctoerofu.30.ion the thorional.).mleit r oh't a o

of Figue 19w proviade tote leans of spcumlaxtvey onstraitingathon leve

oFo reerh co mtstsoe that o migh peeexpeted if the Deprten ofl~ Defenswieto pemloye resonalel the astrateg ofgreamc rtion ntsreachi activi-etimes. Assme for Eextpe that anImprovement, fatr fTeractornfabouity 60 ter-scenumtisn desie inpote by toi-poellanta iherdtial guide a M str rominoiere a19,eodprtr th ateced factor, "cmpext Xo 10 No. REets," isveentca

b iur 0 9 proi.e Inwn thaean ns of the currtient denma onstdelivter fore

...............shearc advstance mish evidecd.No lftet uspassumen (vfr cefnser-wvere toM epon oll the svtrate, y we canexpctio ful Isuccessac in2aeceti

lito ouries.Arcume progream.le that is, ofpreveryn foudaceted, afaot l0eart

one wil h. I'owing wotha pa rsIng.fte urn enrto elvr o

t roun.1%, $40 a pound, a re uirement for2 l o e t 1,0 sc n if c r

-4

BEST AVAILABLE COPY

ll) ekI i NIS tfit I ~ pfI~e e WA int.uIarj((. I I I- 01111,1 I

.11 "t 61.10- tha * y t nm 11 it 'l %ie int i n I 4,- 1 upi t,- ! u u' i i.ijvo, I I otIari n It mal d the iw d I it I ilt i. p Ut

I atx kI V 1 . t Itv t 11 i if 1 IV , t 1id uv l I ylit Jn n I .r.u~~ "

2* ~ ~ ~ ~ wil i k l-s Ild oFNS111 rp a the -tilan !1 ig ijlktkr. 0(he 1,yi i1111il"li that, .1fl aga n, li tbo t 1,000 RXO Eveinv t i 1vo ti..stI l.f t,

toab Ih les thaNn3 evhents, 1.utha w1.r1. aci'eal oy idetif ouiedb te'iud

d~e-iIiVteit L a i mchIprovedLif bnc lorasr.onvgaini'et

Workin tte~broth Fiuexmles and usping ; u the y at do imprakei~ I;th pointwhact iske cplatsiie)onal h Smprlvhts InScopehiertilaed eqipment are fafoein ely th to, bein abhovdut moert cot wh E entswedb copItelyet, asi UPJDý-totes lresitrdue than wheEent thawee overal P idpoeentife lby touhIfil

throuhthe rdexmlrae refeinemn cof eachl tcniquer usted Nain thepred

cessor equipment.

These simple cases should demonstrate beyond any reasonable doubtthat'an aggressive, Forwa~rd-looking research strategy with considerableroom for speculative research is essential-if weapon systems are to be

improved on the basis of tieW science and technology gaiied at minimumi

122

B EST AVAI LAB LE CO PY

ha h~, idr~Lit'i ed t I irouqh ay s ming ol (by Iy II~ fitd -r iut ubI "I

I L lif It iiLi I I po I n I Lh V *reI% c1t11 1 n, I "Iv. ce Ih Vr I -v n o , I tfcn i

onl Li;, t iemt, r Ia I 100 tn o a ricatien techniques rcqu1 i ~ toi imp 1 .1 II - il , I ljot

andl I nd ebt o d I ylie' frqec of11 ther-ccrrnc cou. be accelerat1edbyI he ijI IlluI io . ' 1or mofne y in to 6 R1 prograIms. i i t Ih

ha hf uftherl hadllLIf--of th Eventql as, iI.( howcv , Ire cleal AL dv aepIdut ubvnuYothe l~vl .uol n fir recent advvnces in sThnc ortcnooy reryeofant .ciunot;d in

010 t is ,the l-atter g11 arouio ca ienti n techn olS--eurd:og ic llilj'l adhnesthdat pro.a vide th p .fr a veett oers Beause t hbey mutb sequenial i time.Aprnl

ndverised befor anticng belused), itý, isvanot% cvould th~at inrasl ile;aundin wlitOell h significantly ofsenthei realizantio co1advbanccdwerapon.sys thems. so fmoemnyit RDporns

MTreovher, tolfo the e Etents thatecuren manremcenarmy depenest thnotiutherre en may bneune sar dlys in makin rtehnl g -, vresoures (as n abtedi

F Ir tes tin ielasor irou adeoiigf e scientific ord technological advan ae

bilitis Thi, howeer inoles) ithes angmnot ofoalbeud:iadvertie pr fovi e in cargmnt fo adiina fn s oridcaetat inr asddi-funing 1 il Fn siwoldiffirant y patiuaradatae

Next, a oted in he findings (sectzaionon) ofdasinifcan mer

RMorEenoccur, duin syeetentmsa currelont mandgerespnd toy pbeless that

opiriu, hee na b uneesar W ysinmain rsorcs va23bl

fo etn da ri detsn ewsinipco ehooia aa

ItI

If Ii4 t i v I t . A ct v i v I I!-t k1111 &$ir ie 1 11 qI I itIhi 11 titrdm I/ I- Z~ I i fl

.1k~P rd t~l I11111 INJ t1 rellet a I i 1li' I)1114eiitcII-, Wv4I I i p lon t ( it) I Vi I I rcithe ;)V u lllt titl ' 016 thd t 1i .. 11 et 1611 ..1..

011 .t I kit' rIt o aII%%ith budgeto I lilt, o le git (.I f"et '

efir, ~ ~ ~ ~ ~ t thee; 11 y I .C ellIt tVC11hl Ihl iL.1 t I

IIIh~e, he ritrin t(-the dat va suget t any djt' crease iý wolja d bet

ill tirl arederiexpec f tu ed oltha.ti al reasonabrlelti% pece tag Of thevlof identifi~ugraed verse peaiona eurlwls o thet samel sysilt--11tems. fA fetd ki (I-mttI nQt ier foarc these caes lth resegarnchl budgxertoud

th .6ne t ehnology t u of ineduhed systemsexmned Theasnspwned.tO

l;extd austFigueon hw, 33lul a 0 r percent. ofterequeisithe anewsienc

development. sthel brnge foan ccethose retun? depents, upoul inotreasnablyheave beng anticipay~temdebeforeytste Aesuigng began. Ah cosierlabtle pof-weion howtevmr (pehll 20tiu to 3eceta a roughly1-t 1yesrtimae) scould, has

benFireseen andi the nxectessr researcunerakoaben iecntgf thefni a

Welciiith thll bden ofadvanedl mpganaedmerint ftesm systems.~frprcrnweapon-systemondevetopment, and thes casflesnth oredetailhebdgcotract

adtefinitonlongy with insistencaferuo the usigstem ofnfiedried engieloin-meentpcontrat adThe iedfrmntroductione RD iEnvent coulrd nt reasoinably

have beenantiipate beforie sfstem design began.th',Deconsierable' pess

Wit th deto&dacdmngn* ytm- o rcrn

I' 14 II' '.4 * Ii i.~ I t I qII I' .II F 11 I 4 I 1 44 41:, 4 - 4 4. -1 it ' it4

I f t, 14lL' '~ I vu i' t F I4'iitI i.1.44I1

I. Li' I~p1

I4 dj. II',' 1' 1''1141 'IFII 1.. 41 1 t 1 '1.. nFi*It l

I ~ ~ I k4 I' I'1 I IF t14 1.'. I t4 4)! lt4 44'.. !I I I I ii I 'I 41 it f t 4 I

11' It hotib t it Ig VII~ (tV -t .11 If IICI 11- 14' ...If0

k lit J I C44141 V L4hI) I. byr 0., 141I)i.F - II i lit ~I f rt I 411 4iq .j I I I JIt

I.111' inIIq l it. Lt I., a par I I . I ' I I I )1I14. 1.IC ' %4( 11 .1 , 1 IIld i~ .

1L'0C 011n i A'd JiU t l101 i I y Wa';' that. ilie it'sea ich J) roposa i I o rf pti' Iis, ~ )1

litn 'inot hor , mu re poupo I a I.one . The S it Ld tiunl wa-,su 50 ficiuniItl ' dJra -

v.1te thaI theI s reniowned ind ivi dua I found it tiec('s! ary to -, uilIt

svpa rite I0*tory fad1it ic'ls and sock new sourtcLs of run~]ing 1.0 pursuehiF ideas. *l.;Lerejtingl'y. in t) i*east. four examples found durii49 Project111 NDS I[GHT .tLh.'ý new ide cloo velnt u I Iy re su Ited i~n a.. Jonin Fi nnt technolIogy . I fredisonaIb IF o CS L c'f fectLi ven 5sS .cr it r-ia had been enrurced, it iis likelythat those~ now technologies would nlot have be-come ava ilIablIe.

In retro~pedt,. the difficulty of fairly appraising a new science oran untried tF~hnology is apparent. The basis IFo r judging either cost oreffecrilness In those cases-to the extent that the -situations can berecollstruccwd--would not have bea!n adequate to w,,arrant supporting there-;ea rch.

Mluittisource funding is essential to the continuing encouragement of3 l~iewv ideas and the maintenance of viable technology. This. situation is

not compatible with detniled cost-effectiveness previews at senior man-~igement levels. Clearly, in view of the magnitude of the research ef-Forts discuv~sed here, discretionary expenditures. by laboratory- leveliianageriient are essential.

The. second pertinent matter suggested by the HINDSIGHT ddta concernsthe typical amiount of time that elapses betwe~en an idea's conception andt he beg inning of actual research on it. The median delay appea'-s to beafew days. The average delay was 3 to 4+ weeks, a few very extensive

deklays ICCOuintung fortLhe big difforence between mean and miedian times.

On the has it ;hat . thi'. was (44n1y a sam~pli ng of Events . the tent at iv.C~u: io i';that t-ho processes of reseairchl in science and tcechnology

.4r. 'n4r1l' in! ol i ;in t .4 of >(ende~d dIelay. Pirestinmabl1y, in Ivess resou rcesi1r.' quickl Imad.~e iva ilab le, the or igina tor of the idel!a til ms to othier

I2';

I

tt i *w I -I t ~ iil *

)it I tt*fItINY tl 1i ' .* I~~ Ii~~~ IlII v. -. I i~- J

I '). I i I 1 I II I haV~t~ t 11 -'f t vot i11 l it Iýl jiW I i f I I'd .1

jI C.1 h 1 11 H V tfV uLt 1 t llt I 4 lok I I f ltII (ill Y 11 it I' h a v v

bý' th' ; %I iuuly I . li, E IEig.h , a~nd th fit- I till ou1 i~lljdI lylit-1 II untII Id i Ifi I fvi.

t he u Il d ny fig d ) i 'L 1 l~t wd 1 01' '..4t-WA UC I 10 VJdf 'I I I i %ii repurt -,ui p

dot's' IIto I ' ex dtV(d tit I lie operatI j otia I i titl e rp re' I It t itill of I hie ;I, I u~t I nware. t,/

wili h II'lprV tiprocd SysO tilent l ms w rhfyIti oll

Vt01wda191Pyf i 1~oe yojefcieis ijg t:onia ir f!vrl1C~r- A ;ae l L ft i w n437AItig tgi etilvnt:ivnsio ~nuiIy, n tiIiajt . f

%lpetlnt uo te ie nd!p.cccoxitnc o furprmtr 126or.

'Ilerec-NniLio of ice, asoure o ides in te fom o -ineduate

P! t0 ll 1: lxI.0!!it Ici Jijn~ Il 1 il b 'q.i oqm I~ I - d I it',I hIv 1 r .01 i vAI., I

C I~iI~ It Lyl , I.I Ic Iy tI I nucj I li Intu .I 1 11.it od byi l l I It. pel,1% s ( .111 1 i i "qiIiu"by' in 4,11 I ted itu It' II.- I onI I IU IIt*IT '1 numIber 0ri Iue 111 . I vu. IterA 1 1 fI In e ii :it oin v~ I bo ut reilpt I r I I.It Ic i I I d who ke~sc of t lie

'ItI Ihd Iei A A t t hI Iae ime , par cI I ats in th stu I Ishv

i Ilk V.,11)~ sIst' .11c t .)I wea Knes ues Lnd t hose 5 , tha ouid bt mosr;awIt c chnoI 1: f i .nt

Evied nd thug i :isiý c Iabu realna' sucinc d fincitio o akne RX Event wis

reahdilty. .pricAte theate thr may be seveiprals Even ts in asientific orshe- ccncrewihmtoologicaladac tha I coorigh d clonsdr reatreasentsc (tonlysbingle1 Even,'t. Thijehdooia weaknesse n[toL was ouIrcumvend toos someifian

extent by using teamstof , 10ors 20lf.re investigtr nasigeytmV and n thihpovide an rveragingbl effeict.dfnto fa DEetwsMorbised imthnt invsthegacto wtha the etotalnsitudyisi rexpativelycin

senstieretoan the exact noulbrovnsfud idniymrEvntstuyn ae w iuen asyomorraIn lye whpreconlyaoe thtwoo thrray esevea Events wer idscentified thovlu-ehof newa rsarcidiangs to weapon systems might: bosier repreet; orl tasvidnle could. bei mnepeteodtolo lean that cvirycumvelle bto sharply

-................. ruextn re seirc tefort migh suff20icvesfrthegfturs. In 'thnge range of

* ~ ~ I .~e 20tohered onyoe foo he Events hoepriualwhere adegreadth rvarilue

Of the utilized areas of a -technological or SCientifiL discipline~ isnoted, the evidence clearly de'nies the potential of a lmited resoiirchp program. *

127

!ic;Ittet i H ,,tllt'Vl l lirt frti th i he i it t ,i t- -I Ilh I a- .hIt.t

.,-Pt "i " ed -,ii•- II~ ll n lh, 1'H .'• -1' i ti .,I ,I , Iw l• I e I .,.,,

It t 0 -vrne. , t .Vm I t ili .

"i Jont iI i lk t I Ut .I the that tIt .IIjtI tt lt itt, 1t fttvIVA I I .r ,4I i'Ir I itVI4 II 2 h for the. t-;41l i•1 i t lt u'Iy. h 114tltit0lil• .ii may lIt:

t Ised ,nWith 3 fIt hit lANC .liit- i- ,IhIt., I I a t Ior Ili Mark $6 mi r, ,n t-i

y .2The Experieirni 's R.peatability 1

t A e ameplis g of all RXt I Eveitits, rather than an exhaustive ,tudy,

tends to mitigatle against the expectancy ofa high d1egree of ixperiment..repeatability. Tht is, ifa differenttea I were to repe6at, the study of

" a system--particularly oXe like the C-lah, in which the sample studied

constituted about 20 percent of the Events identified-a different set"of Events might be found. Conceivably, that could introduce marked

A differences in the distribution of factors discussed in section 4,. '"Principal Findings."

STo-measure the probability that different groups of technicallycompetent investigators would identify significantly different Events"as most important to a given system, two separate studies of the Mark 46Torpedn were made. One team investigated almost 40 percent m ore Events"than the other. Although the larger sample still contained less than anestimated 70 percent of the possible Events, It included approximately"90 percent of the Events studied by the other team.

More important, however, distributions among such matters- as fundingsources-, research-performing'agencles, Event costs,- etc., were essential-ly the same. In retrospect, this is not surprising, for, without regard.to the weapon systems studied, these distributions-remained sensiblyconstant during the time the HINDSIGHT data base grew from 100 Events"drawn from seven weapon systems to 710 Events from 20 systems.

8.3 Basis for Invention Claims

"I" Another question on the study methodology concerned the teams'

apparent willingness to accept claims for inventions in the abse-nce ofpatents or other documentary evidence. In terms of the Project's objec-tives, this is not a weakness; and it, some cases the investigators infact doubted the absolute validity of claims. If the identified groupwas really not the original inventor but honestly, thought it was, as longas management agreed with the performing group, the environment in which

1>1 8 i 128

AA

siI IIss 1/s si hoiIi I I t Itt L it stI siI I I o. 11 ý'hut Is I I si ji in

s i Io il d i r 1,; 1 tsd ill 1 1 1 i In 6. f II, X . !1 4 ! 1- .)i

di -, tiell 1" s-, i It. rt%'sit Id %I t 11h 111 Yq S ilce, II)t 1 1 lii C tL~ Iu ruec )I I l NI I t If,r wa-5~~~~~v till Lt]e I0I a I ...k.ioi sacnwege stjl t~c~u n

t A .- J I i e in e tha oil 6, i.. rat,~ 1d 1 ted L: a in ofi NDS ..r," 111 . -iithýeo -J1c Livenl. Sutumiry , i rvle I udedb dering t. rvliu t exten -Y"t vil ap I rritan.c-

vances fin science ond technology. ThereforL, no valid conclusion', Lziflbe drawn fromt HINDS IGHT concerning tlre long-termt value of undi rectedlbasic research, since many of these important events (all out-side thut,tiant, frame off thiaýs study, (194.6-1963).

-8.5 investigation of Failures

Project HI.NDSIGHT'.s methodology mnight -also be criticized because nocon trolI g roup has been establ1i shed; ,but, fa ilIurp~s are. ext rcmcly d iff~iculItto define, and failures or unutilized RXD Events have not been inv~sti-gated.

For instance,, although it never became operational, the NAVAHOmissi le was a most pro~lific. source-of useful technology. :It appears,therefore, that profitable lessons regardinug poor rese-rch managementcannot be learned from a study of unsuccessful weapon systems.. -Research

? is undertaken in the.-quest for knowledge, and the disproof of one -hy-pothesis may be just as important as the proof of another. :Thus perhaps

only inconclusive research ma~y be classed as a failure. *But the diffi-I ~culty of discriminating betweetn inconclusive and Incomplete research isa delicate prob~lem that is well beyond the scope of the teams identify-ing RXD Events, ..-

1929

112

:4j.~~vi ~"t' -IlIhl Iv in AIILI IliI I dlnti~l~(I.1~,Ideu l~~

I' tei .ent I ii1 o e t si~ x. ssi N 'ip It I~ til %.1 1 i~t itI II 1- 1 h i I -t .~i

t UA ) ao, qi l- Sj IIAr it'll I i AI evu kl I[h I as'.iii I i ItIIisI P1IIIit,

Mor %u I I tj Ie t~l I I I i "aii n IU t A it CI lIf I titfi i -111 1l . lie 1 A~if ~I

At' i if II lIvI_ ti olionI ' Ivt( i it'y'

Desp. il't) t he vo-tillill oI kil in I senaI ior In ur Defcnse prrcI IXdto fden t 0yelI 1 IX IvnS othere wa ca nIcIIfit jovrrk thte pus ih-t 1I t liy I I ht

4%.tv t a t persona 1.1itanter lles rt in th ie I DoID laoaoie arI -Iv il- m igh t tt be eet"'

unoubftedlyl' o tildo toncesthe newumbevltcitiitral- l1 of E ve Idetihfed socr

rng in the oD ltiaboratores in rtvio tenor inho use orignatng, ipth- rtnliver-

Ao id untif rbalncings teffec was notedrnndvdrscribed, however, byf*bime

ofithe ea peonisinterest ijy them. In theorwordes oft mihe bea cxpeteido

the MINUTEMAN 11 study:

It must be realized that industry is much more ready~toanswer government solicitations for reports of cont~ributoryresearch than even government labs oruniversities would bebiecause of the profit motive. Industry was,: in most cases,.happy -to supply our team wi th reports'to show, the!ir researchcontributions to MINUTEMAN; whereas government labs had oft~enchanged personnel and did not have either the time, the his-ýtori cal records, or as great ;I motivation to show- the ir rei-lotoand indirect contribution to MINUTEMAN, Industry', contribit-t ion was more di rect ly app I(,( )illan more %easily and quick lyobtained from records.

* ~130 F

JC Ii~ t'Itr I d' tI I tid III c ve t'It',

I.t Di ,Isu i i ja l Evtmt 1 Iyp dc's-ti 1Il-ol. tI Il- II.

d I' tit i ho i on tlsi qt. i it .p r ps b i mi nptd injI I y Iy asyttpt ifl VI 1 ap-prild ting I cut tultn. I ,I A i t Ithus, duringIi. 1kId -thi 1type of study I ltyp IV ,ica ,:l i

dcompIonen1IIt , TYhesel 1,e0re trid tt flicas I"e wlme ts olil sor- of1 th ty I

PUI- IVI CIIIi( OI X) Spcific craive ' civity1 in desIgI n ~ 'l egn rin tha Qa1.hth e d fet of advanc I i i t h p(1- tat1k. fe of the art i s a fere qunl patenidc ffi11 L I:.tab le,

andis esntial' ( o I etn spciictis Suc Vctivity ha een

ciDd wit -huion thecls' o RXDn Events nsieoitadkrcrce,

since 1)I there sem s t so be niotgod reawbon for ludi ngprac it ,al fieiI. (3)nt e Explsoratoryi c deveompmen in smaerkn L an oerd m ateur.(ialsdp Ir csi b nq whri tchi pischaracte.eristically prormed b)y indsymtr;C y. ap

* (4)ili O-11lther11S -xplorato ry n dvlpeti alseof ptedoy, antypclinduerinmayl ea o w

(5 Rniesearchg performedin induotria or govterme nt labor-

Lore. efet fnvacn the Untestates oaboad theart, lid direcuetly to exploatory

cldede wio et..i h ls fRDEet nsieo t ~ .:?caatr

ri(3) Explo~resatryh inunveropmitie in m t oral andp matra i alry

prccesiq, hih s carctrisicll pefome byinusry

(4 te xloaoydvlpmnas edm a

cm i --trw

tI~~ tq~ hv' ma hea a4l In i I i

I .at t qIm .t ed Cit ct~t' i aI c"an h nil at t fI -ul Ia' q vil -tt u a I i iu'll 1 IN If) rt ( ti if rl I v 111011 c t~ of It 111 I c )11 ww I 11V I Gq

;ip

I a ~ 1 I I 1 6'

NSF 06-21p, 19.66, 15).

CiIkIJohn S. , Gould, Willii am S., .- I a. Ti,,I.L.l:J 2

Denver v, Colorado- . tnnvr. Rescarch Inst i Lute, NASA CR-790, June 1961.

" How Many Eng inee rs?"' I~i ~ n, ?wqt'. -In1.int No. .5, Eng inee rs'JuintI Counc il, New.York, July 1966.

L iebe rmarin H.R. "TechnolIogy Gap, Upsets Eu rope,' Nvo Yorak Tbwar 12March 1967.

AriLhuf D. Little,~ Inc.. Springfield, Virginia: Clearinghouse forAFederal Scientific and Technical Information., AD-,618 321, April1 1965.

Marquis, D.G. , and Allen, T.J. Amnarloan Paychoioqiat, 21, 1052, 1966.

*Meyerson, M. 'Price of Admission into the Defense Business," llnv'ir4v)eHvi.:.j ..c'~iw. 45, 4,' 1967.

k -ZA 0J. 3!i0uV01AY Ln Ila-iowze and Technology,. Washington, D.C.:

National Scie~nce Foundation, NSF 63-23, 1963.

'.* . *toa.:ývr. Menlo Park, California: Stanford Research Institute,

1965-1966.

"Resources of Scicntific and Technical Personnel in Lhv. OECD Area, 1963,"1':ii:.% 1 0.dM-E4 Paris: Ornanization for Economic

Coop rat ion and IDeve Ilopnen t, 1965.

.............. .."I.'M Ho:~j~:r I';.ourtwonn. Wash ing ton, D.C. : Nat ional1 ScienceF-.ijnd;,t i on,. NSF G65-11 , May 1965; ond NSF 66-33, December 1966.

133

.A.

Si-h n loon %t Ill'in n' I~ If ma alvu(1 f* tu . (tit,-

utlivet", i

-A ?' ''ioltto O.C.: Wit U*.il *VI'i***

~:~ li~lo ~Menlo Park, Cal ifornia: StanfordRLosoarc l Itst-itute, NOvtember 1965.

or e'-!L 'P Washington, D.C. : Office of the Direclor ofDefense Research and Engineering, 15 December 1964.

Warkov * S., and M1arsh , J. Tho lA'ducct ion wid Tiwiniiuj of Am'~a:~ m., Knybg-"W0Al: I6.Chicago: National Opinion Research

Center, University of Chicago, October 1965.

114

-AA-

1i~10 M 1DI II I W- 1F.1 Ak' I i AM , F N *r'.F I lot',.

FOR A.1: 1: 1.': !'.\X I

A*' 1: ANT ~I:I r r.\NY 01- 111F NAVY I iI):U~'~'I1V Ay oW T11FF AllF ''W FIH

P".pill S ud Pt1101li'dv-1h al.,d or ll how grl p fil il-udT O1%-t11o 19,1 Iolmps t it, sIro v rg , r th1eu F.Iyj)OtFruivis folde r coi 1art

I,, rt lil uat t 1-1. it Le, IIIL'aur oI Julio 190,i). t-o iy' off to wtiipu syvt. ems of

(Ro.port N.11528,i Wk-17 isn 19b5)d has( qti, dvIbot1h% the efi. ie of1111 SVITIana1'tg10 li) tan over11 1 allt pofft or iit.heon cfenue -ache part od exIIho )

b.1pLieve to i 110 assoe MIds com theeoe 1o 0101e onV.I hilorlv a i

t'omrc icntiLye sanaLishs or tohe ipac dspofu th yptesearc madd tecno fgn ar

subslas of qa Inuithe rv iez of Liprath waon vea paoins tow in IPI use, um 1n YrcuLi, Doarmnt ofii advansed staes mofi eg rneserch ad velopment. 0gy stud

w~ilally suili (a)rt tofiient diy House foirmlyu ont.ai l Duinnf AppopriationsNo.''r i id d n28 1 Jun .19&5 prog ams qhIcilw I( na uesindbt t at efi'dwIF lic of~

mati 'aF nit nd rI titoera~ll payodfce oy thesDefe. SIcIieniis (F)) tof miisi

LI,,' rjt'r;sF! ztialsis (if cstile. impct of reseaFrch and cretodi'nio rat ion oF

-,':I; or vn Lits eladvlipa , d stoage FrPodtces of0 vgwriii (wien s vije Tcnhi's F cut i -

f 'i o) Wl i cli i a; ass- F~igab Ic I.,Li any 1),rE of Ltt, Iota! DOD[v~cina inru-ý;r-:orc, inud I kecliciI ogy It k v urrenti F y est hilnl-e ~it ti*t -otui t iIn AlI

1 T)

I l~ I o i .Alt I * fl1lU l .0 . -t , itlU . iI1 I'. I

IvCI I 11V h I si . aIe

~'V:llI '11l its, lhog j1V hetp I itl Io --I Ii lIlCIaI Ip o t op l

.':t 11.,wi1 Ik xI1it:ge Iti t I1 0 1 ...1 wl ii- I ti t I 1 ex el. is r1 : t'.se.11 it I l-:.Io eeIid. oi~ a'Io ra . I 011 I pe v I ; II x'Ii e I

I itIs t~k 11k' iev to.' 11lIOII~t'di -1 111ik Isl y ol1ro it 1 ril~ 1.i

a hu' I e job than coli r~itor rgani ?.t 111111I'tllI O i.1 ) 11111 y Ii-4-1rore t..v i t Wpe I'i' t I h) UI1 I Iat II the t .I It l~rli!i1l 1111 1 -1, 1'O Al e I i I

SI'. almý i :.we Mxsi L..'IIC at ii v i Dellar t~t i I I'nd Is plrpclll liy v 1iA.' I (I I heIly. 110.. 1L'i00~~ to . 1h 1 v cesIr iI (ormot io us iii of w i Ill-

he"* I'dpri..'Lr1 andl phii thll I ti uen f t I tonalivvio of iplr ii' ipa nn wll

Iu I at mta II. ih tilas pon is. Iatcua Imotn siceOico (

I w ill ii apreiate you~tra giving. high;.-. w pririy. Iftov reqet ifor peoleitt..forgao.the types; a tof l upport. Subs111til Intm h zieial participatio ofiom of-pyourmost alleW and expe.Irienced peitple is esiel tir l to li xete i ::v is ofse ,thi .

thisjLct. ISpeiic equests witllots be( dipeciad toiyontrou-tghoip wil douta bettearc and) Tecnoog. ein c~uritrognzturn, weioll e o:in) t lohey wllt both,yore vn reprereaivenon (h) ustring th ometittee wh fring hvals peenl woithrnspnda udi:g the pilotiv prgams bnandli thio w~rivll contnueits (v stiheyhole mor te, rew project.- oLf eusr noma-ol uhO hc l

liv phropdetails cand bet ironedoutcattionle Raluearch pand a ngineeril IgPetiiy counci Lt ein oI Jy1. Th oweveit I priclr, ifmyortn disrev offtho thivpeneraly giap~fteprojl laecntif Is tofomuldate e.o impre ove poth way of

thethkagn creserchs and tehnoog prcrearyoms. setm ear etn

Ias wiltl appiy hate yourgiionsconsequenoeitt toheques- frorm paegoplries

motsetmabe lna e: rete epl sesnilLtCiesie ftipojetlinpecofi Pl quens forb irce oyo -ru yDptforo Reeticarch ndiTehooyHeintr,-ilbewkngcslyWh

yor epesn~tie n h seein cmaitee wih asben o136irn

an udn h io roriAadwihwllcniu i iia

roefrtenwpoet

.\t0i.'Iii I I

I s, I v*i 1)~' pi I mol Ii. oiw Iti Ill I~i I i' : i:, w; "al 1, ag i ~.Ii I I(a,

F ;1011[1 '1 .1 it't- Ll l wItt. tiltl o : t ii i : I . I' vcit 1 it c Ia I c liptI at c. i as. Weii. I

II I ' ar ct..ltds. witl th I e siiippiilI if "thi.' Ansski mI in S't'tI.'t ar. (Ib ) o! Llit,Si I liI. wv D"111 tI- ITIslit W1 lit WhilCh 11h1W IL-l cain nwiiilwt hs. Vs;cI d

Ilk- 0111 i i.L* gokiop tcilti fa 1 3 'ldiv pl ann ing, sesion in tin. PentalgultI 0s'aui :11 ,st .1 il . L110. (1i -0c Lar, and tin' stecriilug conm ilLIt-C will1 coill- nil.l

to' .:s's'- a - r.gt at ' rg iar Itilrvai s thirouigloulL the. st~ud-'

.1.j ins' - Ntivember- 1I

Thei. qLtidi,,!-- of till flirst group of weapon01 svstems ure performt-id and'otnp, I o.'I vd . liii' ob.1c' t (ve. is to idontify and doescribe the RXD events,

cltruint tlii. p lti.' and t tow of origin, and certain other factors in-C1lini g voelctetd aspecLs vf tile WLinaguieinLnt environment in the originating

Noev.nllwr I - Januarv 3.

Repacmet ta-v o [ -me mberts each are f ormvd! around a ie'onld

!4ti~t r,r svt:Letnss The captains of thle nIOW tcavs will either be formecr*L.C.L~~i'-; r bec recruited from former team mieniherit. A SeCond groupk ~ ~ ~ e -oo's'isp' s's~ aro studiled.

i~tt~I .i.. irci' formsd mtitond It 1h.i rd proup of sstr:

137

A A ~ral i :mitt v m - .iv o tt1 t1. i I I I u--

It lit .1 Itll

I .w~ ti Lac i v h t t-atmk; tgt l nt, Ill.t avterang attlm V.'u otmte !.Vý I eni p-tmoll i e. l ital Iltmn 'i n i Ii' 01 ~ I i t-~ !;I~ e ) , I h it, I gmt. 1i I I I ve-

hiel't ettv. re n tIll he In mnt It A. I vve I utly itt t m va I t oul !;e-tI vribie IM.1 reh I .This w! 1Is I -I l I' li. on; It'l .it 1 I v;! I henl htaIlI I he ojVetI 'J 7I t tilt ti t i'e Ws 111I * tk 1p .110 I enet' h ,m I';ltee ga i t met!, ic ' ... jet nt Iw he .I= tilt vini I hev 1as ro I w Ittnlt hIt lly tit for it a i on il nt itiot eI*mn Ivan';.1hi-%e do' inmttiond aft ert t lit I i rn I L irtre I vills 1 av, 4t041p In t'I edI he- I wol It

riheMii.'1Li La my Ile Ia r I ln lt S w I I 1 11r v I,41 tie t vallutp nin annI Iett vail "l-mltthm

WIit (11 e1intla i. I. ravel funds antI o~lier supportl ic ti d I ngy of I I re spact annt

41

I.i I~\i vc.'.li ~I I I x 0 1 iA I ,I dlcv"'IciA.cIcl I I. V,'ic ''I I II- .Il cI IIII

whiic i A. t h ji Al. % It; C'X.Iccc11 i iII 4,ic tA-.cI. TIi. I, I lX I v t I cI I (.i I I (I:. Ii 1 1:. 1'

I pi itl I a v sci'i ii -;1. Acr xccIuacj'l c ':-jIcram on

W4 hicti i ii E .t di I I t i tl l ii 11 . t.II. ickucd o a r Icci lVi L V an t Ic , 111; , Lii.

i.voilt --i.s t V 1icv;Ii Iv I c ,it a rc~t I . i vcv aI ccaorI I t lii 1 ii (Iic I rccili ;i f.'~ -w~ ..al Ic'W [liut' I Is FAAI' t X~. IIIII) I v At nouiic I i 11 lit t lit- paz 'I , I hui I cI'.;i or-' lill) .,I

A) ALI Ild I V IcIl~l I Lil- Wcc k, )' I Iil~l I ha I 1 it cV(Idcltvlailil i (-Ic I I V I-.0hic'rat --Ii f :;

.LII1.11 htz 'rtg oI cl' ' ' 10I oc I.! or It, ft.c r co l mpru.ci aroicinliarat. t V r i st ic' kcs oor t1 lcit I. otI.or ic'ar iiigsi cu rnc- 'it IV avai ai 1 o01 , T. A.i

.1 C'ivt 01 V iI t'' V. of or' I'l V Even t Wo, CUM; 1i 'oi i St oil I Y (f -ioi Itiflic '1 ill-.'ccIIIV A.'d 0 i ( hor I. it iiorot- i ca I Oxn I irlnal i oil o "I I Ic,'e-ii fl Ie u t 'nvv iw harihgjg041c'i10Etl~ 0 5t ttV L I V~ -rd ci n 5I I oll-C r-y 1 il -g e ipae i ty or ;I ,juimple cilt' r!tu; -

s ~ ~ ~ ~ ~~~l .1 r:)I1itt, 007iik a oldh's~pc tvtd on ;I Iydrudyliami i c.ll11

i'ýc'norj iz. I bSear tg.01 jo dI lin othevr :worclr4 , OnIA lyit' Il in ct i gal'U i ()I ;lt*lcquahce to demloustrati.' teI'IICMnit fuaS~ib iti.ty Of Oluimolcndaincc'c-

t -Il . [ddOi i s91 ine [oied. It is 1imporltant Lo note that ii I S EXV 11 L dot's notoxtc.'td to a pro0to tyle, gyro develIopmnent.

S pet' lit ConIs 1 de I-t ions..

Fo r purposes of the Deparuncnit of DMCIfcn 8L'sudy oIF RXI) (lifeýtiVo-U11SISP th~ose EIc2Llts that contributed to developing a' weapon or othermi.lILUnry sys Lem ol a.'i mi~litary capability are of *primary interust. Ano i uonI:r ib tit ing Event Might Ile of intlerest, if the idea has been Leu;Lcd

and is recogniacciz as contincinp, t6 be of a high potenitial value I)uL, fuor

soine tuechnical reason, has not yet been explioi ted.

tFinally, the matter of idea novelty waifrants clarification. TlirRXi) tf~ectivenLwiss study is addressed to the Aiscertainine of CUnviron-menata factors associated urith tile successful prosk-cution anidi t!it 1 izaL i oil

of resuilUs of XIM). It can be safely assumed that a given 1dea~ W~OUld gettih't, 11 treatment Wiietlcer .'or* not it was.* iln fact novel -r5 long, Il-; tiicvoworkers and soipurvisors cif the i.,nnovator thought it to be urovel.1.T

i: r ,r Nio for' rhovell~y, Lhn ssolely thle Cotemporary c2Ifi1iicll 6f tueLmihl~lL' di In to sponlsors of Ltie ideut.

TI)tLeI fiic'i A-t~ivit

Based on tiret abvede IF ii It ion of an RXI) Event , tir hI siri' f h

dItvel opflt'nt. of ;I ne0w dvi cc or corupoielilt can be expoeted to conita in I

139

F Activity

l't- ti* I' Ilit'tttt'tC tf ITime

Pril.Ii. tle leul-elet.ofOWit i de (at illr I juni ufIim rat iti,

lt'e illutovil or Was invol wed inllfli1w :u'qii i.- it I on uf know I tiljg' , tirii rough[itt :i u dnt a I Ily watt u ptrt itit-nl to I'lie

Eacwh u I t~le minort t'xeorni lon; to hte, 1eft of I indivtateS O '.t' (Iq lit I s-j n

of it bit of kntitwiedge. ThOse bIt. t Cons 1St ul' U-ltehtiil1 know-how. iip-r-at Lug enwivronmnional faiet~or:, and., pe'rhapis, opt'rat loona rv'qoi retiun't n. At

t. -wtine t rtggt'ring, elemlent. enath td. tile culminlating of all of lb'- pri---eeusi toite Ldez.. Suibsequient.ly, tile idea In sho~wn 'to he v'niid'r are-

t [VV invest f gat tion.

Cout. innuing with the gene;ýrali'zat [ot, assm titat the output of tile'above Event is in some manner related to another innovator. 'If Lthe

latter individual makes use of the hiformatton ina th~e developing i

another idea, the information concerning thle first Event becomes- one ofthe minor .vxt'ursions in the latter's. knowledge-accrual diagram,' a-, inFigure 2.

A

At

t 414

*1 y4$ Time

''.1p '.; o It I It I .11 I i 'I I 1. I I ; I t11ovi I .1l LI: 1*i I-o -iI)L

Iat- 11 I I i tl t

WIN

j:.Preceding Page Blank

R.,d f lt vei' t' .1) fl

I t d .z-1 1 I ,V l,, ti, a -s V ( tv t ~i *oa d'l:ituns rat ioen, I uivest I 1- ;it i on * 1 idy , vie.) wh ich~ u 1a 11tiluati ed ill

WhI0b--lAtl .u lt' t I u p et r olstela ,dvilons~t rait hat if' pr i nei pies.. OF Npvc I It0emb1d Wtnt ot, prt ivneiples (g.*teubeitnque, devvire, inat vr ia. v tc.)

tN et e Aa RXI) IEvvt I, eottecived here a-, orrespund fuf' to at periotdt, t t ecliitu ai act iviy lvWifth a well -do f i ted end ploint ( Oig ,te prepara-I Lon of it rL-port , prtoneiitaL fun of 'a technicla 1 paper at at professional

sce meet tug , patent. discloaure, demonstrat Ion of relts i ity rfade by miathlematival atutfyslit or breadboard or hrlisslnntrd niodvi , le.)

II~ f Ycal ly a crtedliv e or Innovativ lvaet Is involved. Care% should lietatkenl to avoid (a) fttciuslofn of niormual engineering aetivity within the

-tejwrary state of the art, (b) lumping a nutaber of I4XIJ -.Ii s ntanI I 1-(IV nit class of such activity, and (e) confusing manufacturedhardware with RXD LventsA.

2.Weatpon s;ystem

Namo, Including the st.andard nomeonclature and the common name ifAtihleod for eaqv identification.

3. stlbsvstem

Re f erence to ;tn analysis of the weapon systalm into immediate andseparatelyv ident f Fiaie const ituenuts, arbitrarily adopted as standar.-I

For pirposes of' this study.

[Note. For this. prpose, overall Syptm Concept, AerodynamiteColf igurat fun, etc., willf he treated ats subsys tens where considere.d

114 Iervoree to tau analysis of Ole subsystemi into mated Etat e anldvptt,1y vI dent Ifi thb v ompsnu'nts , cons;i dered at:; N ve lvi ug RXI) E vents

No :For thIis jirpose , the Nbithyst em iem eept , or thfe sunbsystemu- it-Finvdl ihndaovv, will Ift beIrvated ats at) eivlemnt Whetre eunsid.4-

.4 BEST AM LtAB LE COPY

4 . j! lit I,; I.

A0 I It I ti. t11. I I ,I lI I I; t P. I I

in t I 1-1 1 ': 1 a 1,111 t I1 v. Wp .I. i

A Iliov t atc o lactiL' v II iAvi .1(i4 . I t' -It- r ig vf 1:1. It r o ui ii Lt ~1I II Wi t Ves ut .I di-c I 11it I it 1.,, 1.1 vl 1,tI on '11 ro(t'vI )plte, 1~I ma I ItI;I., It i t! p-

1.rkwt,!h ý (Jovc 1W 1 241011 t 1)11 VII Otl 1in .s (1i ~111101(1OlI_! t to Di ,to p a of Ii) r 01 ThoL o r ph rp s..Iti6 o1 ' 1 I ae t~ I v [dual r� i I Ion n. ;I-;-;ei 11 .'. l g Ito

woth Event brior dupt ip Uil form.. 'Ihe keip'ron I' oe biaclhe utnl od .11 II~p*

thev o rganiz'at ion whore dihe RX1) was performed'Il .il ;I a goverInment projectti1ffe (it- ora bortLutrv, *or ol24owhere.

MILO of E.vent

T ilh erinwthte spiecific 1W)D IEvent activity terrminatud (,Ice

I abvq) A, aredetail ed spocif Leatio of the data' shoul cl e ncl uded'i

The ppro imate l.Cugtlh (s;) r1il t ine coveroil by thle spec IfiI t (-lia i 'a I

xot: v! y ba Ingth teorm mavt loll ill 8 above.

I441

LI

lb.'.. .: . . . d aatl lt i aa ta aa .... 'ali.. .... .at~a . ' Iia

t. '.' t l~i' 'Yx iv nn ): a.'i i,,L,.l ;'a'tli J.iaala'd ad aa:! l laa ad l a''

I k.

j�:-'"a': I~i g.I I "' it t it l : til ,I I1 -I h i .a I Ia l!A•.I I." 'it - ;it. ja il

-4,'.)t .it I ti la't l .1 , tr •iotl.t 1- at d t, l .I i 1i , t" t ,t I Al:a tl I I v IIlata tI

"- alt rwii r Itc' lJIM 11f, org t Ih-of jIll) i.'r s t i-a a.

s Noni ta 01 s,1 Io;lit r1 i-1 1en docrip 1i hatera :to I larify In]ix Iti

"- I o"titit tia , "1I%. ) I (p it) ol laltoryII, , I) In Ivsral pro! itJ.• -tit imp•)o rt tit pa I' I in I 'I-- RXI) I:.vt-n 1 ,id • ili d -Io•O I !;t, hit- H o wq~ il itA w ltu rt.

; .: 2 - I 1.. {t rill, r a I v ,• l'J]

• : .Tho ttt'nerlt" |tyI}s of o-r~atilzat ir)y vo'rrel)ondilij, to It) ahuovt. TIhII!;2•.•. ~shouhld Contain) taifli.cleut: dt, scriptive wiatterfal toc (larify fully the,";•-•'• ~~~tIW opsf orga,,iv:at i,, a,,d orga,,izatio,,al subdivi sio, n litqust ion, eg.,Sirndiit|riaI (prol it)-tcorporai~e-re.sea;rchi lab~oratory, Industrial- (prof it})'

--,t1i-t',' ing divis ion-ides Ign engiLnccring organizationi, uhiversity-

"operated Department of Defense research laboratory, etc. Thie purpose, of "thils paragraph (see' also 6 above) is to assist in the classification oforgan I zat Ion types.

12. Financial Support

Specific information t)n:

.a. The source(s) of funds. This should include information con-cerning both the internal accounting' treatment of the funds used and the

4. Ultimtate sources of funds. Where the work is sponsored by the govern-ment or other sources external to the organization (10 above), specific,contract or subcontract numbers should be identified where possible.Where the decision is made by the organization (10 above) to initiate

* - hto activity representcd by the RXD Event, the way in which the costsare recovered or treated should be clarified (e.g., the -expression;a"vou..lanv funds" snould refer only to the nonrecovered expenditure of acompavy's earned surplus; where subsequent recovery in the sale ofproda.-t.'- or iF negotiated overhead on government contracts is involved,

i! - n:d ta, so stated).

4fit. The timc duration of each source of funds.

"145

alt,

I I i. ,;;I V I I III. I i 011 AII I ibii .11. I 1I 11I . s i -I 3 3

t:;. 1SI I .IloI II Ii i1 £' %-I ii, hIii I~ Ii i I " ).1 I1.11is :1 11 l 1 1 0,1 - i'1" p

i it%'lI I Idi I I -I .i I p Ir Ih tal wc;hli1 1 i i 1 *1 uhV '0)131lt I t 1 1, sih hi

IX Ilk~l jvli oIi hilihi .131 Ia I IV. pro:V IIIs IM i0 I o 3II W I, 10 if 1-1]: '1 u ii I-s I i

V011III' . 'V [Not lsit I oInforma loll :0,1111 ( 1 3.' ! i II na lin- v I 'u i i; o Lt i.,1- ( II Ii' R Iv Lts ' It' I 'oI I C a,11 n d dil' hII, I~ I i- VOIdt !:'i ~I! I ii its Iarm N, 11!r 1oI ii'

r ti t h* 11 h a 1' .'L 1311 I olag hirt' I-its 1 o i %14.;' Iit'rI o ti lt. Ig 41I) I .a .I 31 ',i lir. l

Liiim ILI IliI' 111 1 .I.*l ('lIII% ~ t ii 1)11V 11 w . .11

a. I I In 1,1 or 1; 1ja-I I Io I'nLrt l k. thii Jay n jwI(I li ) the l XI- V Itwa t1litJ, that11 I,, dh 11' ywihi a iioroae ito nsb

- uct, I r ilelated X) c Vm, nt an r syih stefmst. or imil 2, 3oruid 4 lt~t phovs Wherev 1r 'tI, po~ihlkspc11( in tic eveit s .vohou ld er idtitifiud g g thpoaaill

ZA I. ti ormia t oii conecirt iug priora RXD wvhith sytm labctvity Waori

iai~cdviis' which contributed to, influconcod, or provided a motivation ror,the KXb itl question. 16i pnarticular, where government sponsurshlip of- theLWXI1) work is involved, state. whotlier thec.technicial initiative resided inl

*.the pL'rfsormi~np org~anization, tile government, or els-ewhere.

* 14. RXD Evont Circuti~stultlIes

Mitscellaneous informnationi re.lat Ing to tile MXD Event but not cl se-where classified. :Managemienit-envir'iomential information may be recordedlire Whete.j'ir, because of the nature of the ILXD *Event, it is possibleto Ide .ion Istruite a rdiationlship betwe~en thle cost Of the RXI) Event :and L,110cost saving~s to the goverament of either the final weapon systems or

-~spedl [I fedelivered hardware, thils informlvtoin should be. idiont if-led and.fA

rep~orted here.

1 5. Sources

Wcumnts peronsinterviewvd, etc.

* Aut ho r: TvI'.ep hone ntxm11he rDate of prep. II at i011:

#VALtar lifl.flt : E"Xmiinpli or RXD Event hDcsc ci ptionl

'.140

AtIatf i~t,tI it A p1tvitit

(it [n: I IIIi) I "n;.t 11 ci1. ilo : I I IutI 1. te : '.t L I i I t 4n

,N,.;tvti-: hap• p ier S f i'111 .l I t v I ,I it.-

O v. i gin T'o,"¢ , itt i it -I S 1,At-'t I i y , antI Ot I t'oi'Th,: lVcm , tl t ; I s.i th, C n e tio .a d dt11ls a tt o d te:Ii i

1lh,11 orb i t if a near -t,;it't It sat t' lI iLu on a s I ,.lg e pass by tII accilrativmtn•'..tstt'tittol tf tlth, Doppleor shift pILttern In thet radio IAgnIs•l triams-I I Itt'd by thsa t'lzollitv. Tme lh opplr ti ffecL t. ist anpparetnt chilnge infl'reqnt'ty c Laused by the rel.att ive motion between LrtnsmilLtttr and rei vivyr.

in Re•tatI•nits1lip to Contemporary .Si'nce and Technology.:

Ar the t ime,. of tlits event there were .,eviral inethods of Lraickingsat i ILes. Opt let track Ing Instruments using the visible-light port ionof the eLect romagnet ic spe.ttrum, such as the Baker-Nuan ballistic caltter;awhitch det erntitees angular position by photographing the ve'hiclu agaitnst astlar bia'kground, .arle highly accurate. 'However, their capability is lim-,itead by darkness, clouds and haze-. Also, th e data require specializedhandt ing,: SOtetiieIM,. detaying the output beyond the .period of usefutlness.

Infrared radiation from tlhe satellite permits it to be trackedthrough some hiaze 'conditions, but infrared is absorbed by the lower at-mosphere, limtit:Log its detection in ground stations.

- Radio tracking techniques included radar scanning for tim di-rev'tition of tite strongest signal aad interferometer comparison of signalphases receivedlby separate antennas. Minitrack, which operates fromtI '. latter principle, depends, on a transmitted frequency to establishtie I Ine of positilon between the satellite and Lte trac'king station.

MWti le Minitrack uses a small part of the Doppler shift te utcas-tint miss distances and vannot determine the orbit in) -a single pas-n, th-isevent derives all. six orbit phrnmeters from the total Doppler shift inone pass. Using scalar as opposed to vector measurements allowt'd vast-tsimpLifications whilc retaining accuracy. For example, only small an-tenia•s with a minimum exposure were required. Bumbilebee Series ReportNo. 276 was issued April 1958 and distributed to anyoue interested ittsatellite traeckintg.

i-. 147

">.I4:

n um n Ul ia im • a

(!I,- to W~tviit I h It p l i t,1 Iiv 110 11 1. 1g L I I. ui V ts1 I.. . I I.- III

A. i 1 ItocIctiq tt I .o I'll ti I ilk'a i i t IY I I vIt w 1v1.

* ' tn. Mt~~' ~I) I*XIothI c lvw ito l iIII't-L;i: *H :i;rI It i

t~btC~V~I .t .11'j11 Oil tci k- :o Fi Ltt.11 I l 'to tsIbt IV or i it 7t ; -It (:(t ,

%'I rotV1 o I !0 OVO I Wt 111 Wh t -I I' I :,,it 1 1.1 1 t1 l . 1 o pt , . , ;

.1 i J I i t- i btcI t) I i, Ilot-1)vI Itr ,h ii I (I. tp .1 1x e 1 .1 tot lit ttie t I o v ,.11 I Id IIP1I It Okv I~t It 1lt' ft'I t' hX s I'M 11 I'l ug up t I toi p itoI fr IIo act .

G ~to e I to a v n I.y is fv rout l l i Th ilto I I kip I i'lt of; i I . al ti;v trch (, Itat &I-

181.. D~liuti o f I Io Dllt: r -il l

10. toll ofvnirhutia daon

Dr. Joints NopkItns UnieursityclIh-jt w' vsathCne.A~!scki1h.w iAplietd Pysics froboratorytdottoi:2es.z1 e1,1is

8. Rae o .enict Cne

ua.plied invsiation. p l15

1). Orgnit!at ion Tp: Octobertt 19b57or

1) Durat~on: Six monthsa

1) . Appulied: Phsic im.Laboratory

C.~~ ~ Rep113Cete

I't I i I ~ ~ . II* '*

,: I'~e A.il 'I~ t:

toio m e I Ut I itb t Lie 1 'd .11 I t t I 111 1 1l,:1 .II

IX',~~~~~ II II. t1 I.011 zo I I I tlt l I u n I~ ~ P ..

'. 1 e i lv 001 Ac Ie i v11 t 110, 1t111 vr~iII IIIh ... ll

~ *~-- .1l~X .111.1 eal i..d .11 V..11 I I te I011 I ~ o Il1' I-$ Io a i I

trl II t 1 .' h 'owi'vr Alld ~Iup J1 d I iwr I tsco ot111O 10111i:I'll t i I h~i I oh a I hoil 1.10tt1 oilart OcI I 'll ap I 'I~ 01 I I i nigh Ii Ith I 'l 1,

q1vIne .11 t -va I 01 ud I hato or i II fii mve rol Id ere workea ot. IE " IcIIIatt lit J. i i I dIIII 1.te Iu APllI. a otIOW WUi I Jflinty 1ctuIrI o a I tiv it 1 03I fla ti perm1i altIt Id I ds-~i

Irior 10 hod I no t eitiv i It;ih iostw I 1ipat vomputer I wais i av-ij t I bl Sday illl(I

n1 4il. IM) 1: 1111f Il')'Im I h wredwt tt cmutr ecm vr i

Or.edil taik MtoChnlt. Chaormand creativly- Researchenterl with nomputivaIfntit, pot;d vielatv ions af~iter trao n mothso sateudy, th iti ie Duiorplervlooked(110 m lore nd molie sensitive d it scri00 * aotdnuttirnlifbilty 1in pro-diet~ I~j fgoril wastoI muct greater tan antlicpatid. was nt nigtallyg thist Iint IOrd and opIV t Lik orbiatl ropa rack~eingl u mn were wreOt. i1 a suiedlt -

Vaifidtt-er tire Iavpetrat cmpethod eas tl a-llmwt le ti

pimporlal; itt' oto livel vsilbensiof thile covputwere h wrads aviibloe daatlIl.g1lt or Vnd Wei te.i-11, widl~l) worked Withtit onof tpuer Univate v10eata r I iu-ve,4Ied ill Cie. Dopple lttheory andII c~trliea ytivel keypartimentpd iin ledmaitvr

ltipte andw eacuap erion.e Afteerl w mot-1 oflO sQit! utdaou the iDopjiop Pc.l~aooko nti And. more snswitive nagiend tniqhe, an nvits reliablit irl*a l pre

Sd ioati ori was11 much hgr qester00 tot lnic 1ipatel.d . lvent'i5tley youi tapot-

.. ImpoAl-an to tl'oit, iiarces of th'5 is vvtn werl o Ile iheseartecd Ilel em oytkd

\Ititliith t1)I Ii v Rs~ II ~ o rat oi ohi ownwa tim . it aW 0195 ei vtl gi-od t-ss I U I c

,--varc i at AIT0,i l gav ne . otagm otilt-~nu r 0 11 0l i n I140stjtorwi I )and' listtI ;

iFti~t-fsmftitrq e to s L i utv e . H i- it is tt

-z.,o'AI. c votfv tit iu1ctveo h eeritCne ' ok

11 41 *l v W-i .*I \ac

DW I t 'ilit .l l

Gui 4. Ca d W.I ib tl

L.10 r It rN~ ? Tilt, Aopi ii Ify ~J ill- 8 liin

SAll I''

Altho:1. Vait tL'rsOill .liikt'r, Corp.

TOI I~~o lI.l littirnher: 20:1, S44/i-200')

P.1L 0L) ' 1i 1'01MktilL I Oun I IMay I96

150

APPENI)IX I)

Sunuiary of RXD Lventts

NoLe: Numbers omittedfrom the cons cutive

series represent reportswithdrawn from the study.

1'5

t!"

Bst Available Copy

*11

* GP42

ver

a a

1..1

5~~I do c - :

CS H .0,

Best Available- Copy

42

S. gi ,

~~SL

.0.

pa:

AAc

0C SQ 0

Sit M S 1 u~ r. -- -,

c 71; d

0 -MN. ~0~ ~. ~l*,3

IF~- 5- 5

Best Available Copy

j; -j2

L..* 1. CD

bra 4'

144[ i, L L0 V 1 ..

P". . VI

in m.4 LO c" L

I~71J

3ctAvailable Copy

r ca 0

0. 4,)..j~ fj ID ~N4

kA, in, ,2 7;

2 22tn I Ir

Best AvailableCp

II . 3 ;.a,

.. - to

1.'. 3

Sp

I * ~ ' I "5vs

:5 :

l

Be3--st Available Copy

'low

2t 5

$LJkI

4h

Best Availabl Copy

-I! zi

C-. ij. 11. -1

IIdJ S .' ~ 1

.0 0-.ID

:o 0 0 -I.t 4' ~ 41 *

4, 0o 1, oL4) .. .~ 1 ) ,.,.--

*: i. .. ... cS L 4

ja . " .. - :-- c, ;4

sa-0-0 0

c 55*

0 6 ~ o

A .U~~aW IC.0 ~ -o .4.S C, o. - 4'.. . ..

0~ ~ CD 0 -0

od o 0)(:3u 1

c, u .0A

A 1to4-

41j

o c5

Best Avaiiable COPY

L, 1

T4 -bt;r- oj

0o a) I

V - 4*

tvUgo E -)I

:1 4- a. 1; - * ; . -

c l~~,~ .-j _4 I l 4 n 1 w u I

H 4~ 44 4

Best Available CopyI..I A

. * . .

"a I.a

.1;C••, . .,.:. •- ,

'I I ....

:43I

•! o = • .• It, .

cd CAd '-

' ,• 14 i I "I <I t t iv•, rE 1

*ni i LI

lb. 4)

to .1 -g 0 0 a)&

1) S- C,

u -" -0 oJ

|*!- ao O

ItIa~

I ~ H Ql4 16o

Best Available Copy

412 2

I-I

41 r.0-

'0 V. 'D . F

0. tn -r.

Ito

V1 aI 6 i 6

0 e

A11

1I..

Best Available Copy

S

I I *

g

I

'I

I.

t -5.*�t .45

- . I.. *1.

2. C -� &

-' C' - -

'� -'c 0 -'

J 45

-2 p ci

* CCI[it r -

S-, I

h .5 ')

*5 *2'**

* ii II -.

s 'is

:- .. .

8

S. .. 5..5..

-d

445.

54

r

'0'

a Li

Boat Amvatable Copy

v - L

r. 41ý: f-

-'cst Available COPY

7Y!

-I. I*' 4,

Ij 21

0 .44 & .

CI I

n acic

Bes-t Available Copy

Bv e

-42

A 0

w u

[A j ~ ta Ei F

At

-'-I-

1 6- * ~-

Best Available Copy

Ar

104

Bet Ava~iable COPY

A. t. -m

~~'W cJ J .

-1 N. I . . - . . . n -t ý

Best Available Copy

-5. ,. i ,5.

,* .1 . .. .

. . . Iw "" 4. "' 5 .5.. . :5,5. .. .. .... .* . 3 5. . .. . :. •

.)U. m ... ., i.n , . .. ,,

Si. Ii I: • i: 1 -a in ol I I . ; I..

vi . . ... . ... .... .. .. .. .d

. ..... . . 95 5.,3 5,5:1. 6 8.. C

- 1

""i " U 96 *o* V 4 1 tS. .. jssj - p i- . M 5 .4: 4.5 5.5

-, 4 .I54 i--

±5 4 5 ± 5 ± ± 6 4 5 7 ±.~ 4 56 5 5 555 ss~ 41 457.5 .

'.5 -5 5°J 7. 5 .) 7.5 ( 3 ±35-4 - 5- - -4 - - -4 -5 -

I• -. •., -_z.

... • 5,:

S-", -, • ,.i .• •, •, •n • ". '-: ,- •. 5_-• . -, ,. • .. .. . •J • .. .•. • ',. •• . .. :. ..' :, .•;= •,. _-

I I'•1 45i

�& at

Best Avallablo Copy

It

r*

I,

A. I-

.4 �t

*

I-

A

Ci

-� .

.-4

-

½

2

i: .2

:4

S.

k �.

.4��

-4

u 4

-.

:.

-. .4. .4. -.4 . .4. -.4 � 4. 4.

fl-A

' � N 8

I,

Best Available COPY

MZ

rAf

ork

I I~o4i~va~ab~eCoy

M-IL

Best Available CopyI

II

0 4

1:. . .. - ... ...•.; .

-,• I . " . .' I . . . .. ' ,- ,;

S.... 4.... . . [

2•,.; 5' , , ; i , 4; -

.I• . • ! *.I : .

I 4 0 -.

•5 4 $4 .

"" • - t 4- o -

- -, • 4- 4 **, , *,, ,.$ : 0 :

* /

Best A"vai•"' Copy

"" .. -. . .

I:

A .

.+ +

~

S... ... ..- - - - : . . : .:. L

I * .'

'

+. • L

'A Ia

Best Available Copy jI

VA

A

* z

..........................................................................................

. I ,

.. .

- j .

� 43½

�'1.

Q ¼

It

IS

I.

Best AV a-faF C')py

r:- V. V- '.

It

Best Available Copy

- . - . I,

* .. I

3,

3. :.. .7. -. 3.

* . 3 3 5 . 3.

* 8

.� **7 33 5., (1 1 .55* -7; 1.

3 . 5.

I *., C.

.1 Sr .4 L

* 3.53

'.3

c, � Sc,

-. 3. .3- -73 3, s.

3.1 3.) 33) '5- In 3.. 3.7

c¾.; (3

2

.4 :4 7,3'S

*-. -. -.4.7

'3

3'7,

73 I34 3)

:33

33. .. .7

13 .3) .3.

-l

F * * -

-. . .

-. *. -7 .' *5- ....

I /7.

Besst Availabice Copy

AkI I .

I Best Available Copy

0A

IP:t:*o I-

In In IoIn

M m ' '-

o 0 0 0 ,.

41

.t- .•• • ••' : .

Z.3

.: ..t "

:". .- • "-.r r• ,-t- ?: • • " "

E ''-I..•-:

I, .( -. 5:" " :-v

3)-.,. . ,, I..•.4." . 4

S..tj54' .4 .3, 4/ 4

i"i

, •. ,. •1

'S :) C ) C

Best, Availjable Copy

Ft

FN

It'4r

Best Available Copy

AA.

:II

1; 'n

i 0,

I,4 :VI.

Best Available Cop

0 0C w

.00

0J

> 0

In S 2 In to bl0c 0) 0 4--.

CQ 0 430C

-0 D

000 0k u

C: 0 04 $4 0DoI04. 4

404 0

(41

IL

Ij J

of k. k 5-.g

5 Z 03 u3 : 9 2 3 A 1 3 T

*-0 0~- 0 tu -

41w . o t 0 m0 cc ad. w -.

0 00'

0t v ea s~3a (up .0 -00 0

.00.4 .

5-4 E. 5-44to P4 -

ID

t, V. P5 05 -5 0 5 asa

-0*

,~ ~ ~ 8 00 .lr~24baC

'D a I a 0%

P ro ~=0 . 0 E) Vu 1 W

m2. g54( to.. 0'4 OD m . N~ 0. Ift an

:T. 3~ 3C0 0 C 0>0 0

182

14~

C 13

A4 Ft $4 .$3. $4

00 0D 0 *'

34 as

* be - - p t. 4. **

cl~~~. -0j r :

40 =b404

lz - , a4> 0 I x~ $ $'N I

Oma *

3 j:*~ ~ Itm 4Ž 'ato 2 b

34to

4; 1 4 f0 4 ;10 10tto 43 3 33 .23183

J

SBest Available Copy

"iV, ...•- • - '4 .. .. 3. .,. :' ., . ', , "

'o. "o3 !0

I.3 ."3 3. 3: : • " • ' ' ' • '. ,1 - ": ; . ..,.. .,., 3. . .3.3: .r"33 .* . ... . 33 .3; 333'

-.. 33 .'i ) .3 .-3 .3 .) "". .; .3 . . . 3.

•. .3 3 3. 3.. ,3 .3 .3 .4 .3 ' .3 -.33 34 0

3 2 ... .. 4 .. 33 1 . . ... .. •.I . . . .•

C. 0 3 10 0 03 L 0 3.4 . 0 .)3

C,,

-1 -

4 14 .4 .4 H( H(.( 30 3,3 34 3 3(3 34 3 -1 4 3 33 33 ,3

•.ZV

1- . 1 -0' .3 03 0 33 03 03 0 3 C 3 CC 'C 0' 0 ' go 13 C 0 ' C!,

'184

1, .1 W .) A " # w 1.

I~ .L ...

v r.NI' It DI 4F 4. t j v t , 1r

L 0, 0$, 0

aF , JI 0 , at.. a . a. 01 ON a, . a

-1 . LL. .4 -,,L. a i Ci

w ~ Oft C: 0: Pa 0 1. .,4 a . , a

L.L 1. L LN r IN a In m- N 0 10 N 0 10 . 1 L

'a" D ca0C> C

LL..

:3 14141 cd cc

Of2- 2 wW40

0. .I5 4 4U.ce4- 41

1 4 Lrq . c>zj ~ ~ .g ;.*N0 4 0 4 .0 bb 0 0

to 'a , Z. "

(4 , o

14 4 4).

(4 aQ 4

bL)4In -0 0 . Rm r9. 0 4 4) t;o4 84 v~w.

E44 0 w) w o

4> -. 5 0c> c) . 54

5"~ C5-. 4

(4 E.' ~ .~ 186

in1.

071 - W -9 '

49 a

4 A4 s

0- ~ ~ V ý to U o oI14 ý ' U Ub U

0 5.H0' . M A.

0k I1 14 V4 ta 0.

'Sr S t K .~2 .

f r18O -1

U5- (07. OD 0 0 NI- W1

oa P4 N cq m 04 :

S18

A A ~ 4 4 ~ 4 4 4

CL f

Li 1ý

co t:5

he ~ J N

Z, =

4) 1

0 = 2

o .0 0 0 -0 .

- C ~ - ~ JN Nj.§. >- -, C: :

-2 0 E E

- -0 0 5) B A5

I-s 0i-' Pr

1. V.

Ato

.5 -

ta-bO~ ,,

0I W w: cc ID

A A0

In' -P 4' -P

Q tlt 04 04 Ot 1 0 9

04

P4w~i~

coo to 4

a , ,. -V -0$0 $ 0, w) 0 0 4

.2F u . 0C,

-0on

4.0 4, -gCD Q Cc!zz4'p i0.19

Ar

7 4 b

I _L

0 0 0 o C 01C

z :2 it'

w~ 0) - m ~ -0 t Dm C

O~ 0 c 00 - 0

Iw

00

vs r4 s r a. e - r

0 CD = C> Q C

~191

' (

-po

S...

'4 -- *. -:

r., I S o. .I V

m ~ 0 4co m 10 I L

0 .. 0 t• f I n

$4 24 *I -

s .a": 00

0 10 0 0

.. 0 �- -o ;. c.• - _

0 0 0 0 .2

• ~ * -3 X .a PC , , gg- ....

4:c

A-: 0

0 o

cq 00OD . 0

Go 0 0 Go

IF

Sxt

w C l ta w w V CV

0~ 2.4*12

7 0 70 00 70 V . i

-o m0 .0 3o M

40 .0 10 to 0 r to V) AwV- --

14 34.1~1 ~ 14e . - .

-7 3£~ P P C ' C ' j~-.to

010

w 1-- 14- 14-. 0 C. C.

t .Q'4 -a4o

A4 .A 07 4 0 4)

0 01 -If 0 0 . . 0 0 07 C

-A 0 0 c

j1 024. 4- C4'

* 0

A-~~ -90t. C 1

,C, 21 ý0 > br

.02

rA -V'.-J r-

0 ~§~ ~, *30 4S~~ -a a 2$S93

I. tol 1~

dC* A3

CCiDo - - *

K - - IIU

C1 CC

LN ~ - 4.to toto (

Ea . 4*. 4) I *

x fa

0 -

ACiC

4.) 00 0 *

00 CU I'en 9, -Z , o0 0

d u3 9 -ý 1 .4 -

a) 4) lo > 0 '

4n -P to oU

fn z.~ -r I

~19

Best Available Copy

APPENDIX 1.DOD Fundi_ for RDT&E and Estimated RXD Lxed r

iThe tollowing table shows total Dol) funding for RDT&E from 1947"through 19b3. The estimated RXI) expenditures are based on the assump-t ion that about 20 to 25 percent o0 the RDTHE funds was spent on researchin science and technology.

Total RDT&E Funding Estimated RXD ExpendituresYear($ mill ions) ($ mill ions)

1947 515 .1181948 534 123

1949 608 1401950 539 1241951 758 1741952 1,164 2681953 2,150 4951954 2,187 5051955 2,261 5201956 2,101 4851957 2,406 5551958 2,504 5751959 2,860 6601960 4,710 1,0801961 6,131 1 ,4 101962 7,643 1,308"1963 7,638 1,445

46,709 9,995

195