1991 NASAlASEE SUMMER FACULTY FELLOWSHIP PROGRAM

JOHNF.KENNEDYSPACECENTERUNIVERSITY OF CENTRAL FLORIDA

DEVELOPMENT OF ACOMMON USER INTERFACE FOR THE LAUNCHDECISION SUPPORT SYSTEM

PREPARED BY:

ACADEMIC RANK:

UNIVERSITY AND DEPARTMENT:

NASA/KSC

DIVISION:

BRANCH:

NASA COLLEAGUE:

DATE:

CONTRACT NUMBER:

411

Dr. Jean C. Scholtz

Assistant Professor

Portland State UniversityComputer Science Department

Data Systems

Technical and Information Systems

Arthur E. Beller

August 23, 1991

University of CeQtrai FloridaNASA-NGT-60002 Supplement: 6

https://ntrs.nasa.gov/search.jsp?R=19920010077 2020-07-05T19:45:52+00:00Z

ACKNOWWLEDGEMENTS

I would like to thank Dr. Raymond Hosler of theUniversity of Central Florida, Dr. Mark Beymer and Mr.Dennis Armstrong of the Kennedy Space Center for theirefforts in making the NASA/ASEE Summer Faculty FellowshipProgram an enjoyable and educational summer. Many thank~ ~9

Ms. Kari stiles for answering so many questions prior tostarting the summer program and for her efforts in makingthe program run so smoothly. ~

Special thanks to Art Beller,Greg Hadaller, MarkRicci,and Mike Richardson for all their help with hardwareand software issues. Their input throughout the summer w~s

most valuable. Their efforts at making me feel welcome inthe group were very much appreciated.

Thanks to Frank Merlino and members of the NTDstaff for their assistance. Their input on suggestedrevisions was extremely helpful.

ABSTRACT

The Launch Decision Support System (LDSS) issoftware to be used by the NASA Test Pirector (NTD) in thefiring room during countdown. This software is designed tqassist the NTD with time management, that is, when to resumefrom a hold cO,ndition. This software will assist the NTD ismaking and evaluating alternate plans and will keep himadvised of the existing situation. As such, the interfaceto this software must be designeg to provide the maximumamount of information in the clearest fashion and in atimely manner. This resea~ch involves applying userinterface guidelines to a mature prototype of LDSS anddeveloping displays that will enable the users to easily andefficiently obtain information from the LDSS displays. Thisresearch also extends previou~ work on organizing andprioritizing human-computer interaction knowledge.

412

3

SUMMARY

The Launch Decision Support System (LDSS) is beingdeveloped as an aid to the NASA Test Director (NTD) duringcountdown activities. This report presents suggestedrevisions of many LDSS displays. The revisions weredeveloped by applying human-computer interaction guidelinesto the original interface. Data collected from potentialusers was also considered in developing the revisions. Thisdata was collected via a think aloud protocol, numerousinterviews, and a questionnaire. This data verified thatusers found the color coding of the revisions sUfficient,that they could correctly interpret information coded intoone window rather than three, and that reducing the timelabels was acceptable. The emphasis in developing therevisions was on presenting the information so that it couldbe interpreted quickly, easily and unambiguously by casualusers in a critical, real-time environment.

The task of applying Human-Computer Interactionknowledge is difficult because research done in this area isoften difficult to translate into practical guidelines. Ameans of organizing human-computer interaction knowledgeinto a generic framework is discussed. However this methodalone is not sufficient to be able to use HeI knowledge.Other problems still exist. For example, during therevision process many trade-offs were made when decidingwhich guidelines to apply. A method was used to prioritizeguidelines by examining different characteristics of tasksand users. For various characteristics the criteria ofprimary importance are proposed. Using task and usercharacteristics of LDSS and applying this prioritizingmethod resulted in the following primary criteria:consistency, guidance, workload, and significance of code.

413

section

I

1.11.21.3

II

2.12.22.32.42.52.62.72.8

III

3.13.23.3

IV

TABLE OF CONTENTS

Title

INTRODUCTION

Organiztion of the paperThe Role of the NTDThe Need for Human-Computer InteractionKnowledge·

LOSS REVISED DISPLAYS AND DATA COLLECTIONRESULTS

TMID Revised DisplayFiner Grain TMID DisplayOverview DisplayTest Parameters DisplayWhat If DisplayThe situation Assessment DisplayData CollectionComplexity Issues

TRADE-OFFS IN ORGANIZING HUMAN-COMPUTERINTERACTION KNOWLEDGE

IntroductionOrganization of HeI DataA Method for Prioritizing Trade-offs

CONCLUSIONS\

REFERENCES

414

4

Figure

12345678910111213

141516

ANTDAPUCDTCOBTCOLAGLSHCILCCLDBTLOSSLOXLOX DBLWNTDOISOMITMIDUT

LIST OF ILLUSTRATIONS

Title

Original TMIDRevised TMIDRevised TMID, T-20MRevised TMID, T-20M, Unable to ResumeTMID with fine grain resolutionoverview DisplayDynamic Overview Displayoriginal Test Parameters ScreenRevised Test Parameters Screenoriginal What If DisplayRevised What If Display, Textual VersionRevised What If Display, Graphical VersionRevised What If Display, Graphical showingConstraintsWhat If Default InputRevised situation AssessmentRevised situation Assessment, Display 2

)

LIST OF ABBREVIATIONS AND ACRONYMS

Assistant NASA Test DirectorAuxiliary Power UnitCount Down TimeCrew on Back TimeCollision AvoidanceGround Launch SequencerHuman-Computer InteractionLaunch Commit criteriaLOX (Liquid oxygen) Drain Back TimeLaunch Decision Support SystemLiquid OxygenLiquid Oxygen Drain BackLaunch Window .NASA Test DirectorOperations Information SystemOperations and Maintenance InstructionsTime Management Integrated Displayuniversal Time

415

5

I. INTRODUCTION

1.1 organization of the Paper

This paper documents the application of Human- "computer Interaction knowledge to the interface for the LOSSsoftware for use in the KSC firing room. The first sectiondiscusses the role of the NASA Test Director (NTD) who isseen as the principle user of LDSS. The functionality ofLDSS is explained in this section. section II pre.entsredesigned displays and discusses the rationale behind theredesign. Also included in section II are the results ofdata collection on usage of the displays. A discussion ofcomplexity measures in original and revised displays is alsopresented. section III discusses how human factors "knowledge can be organized and applied to interface designs.A method of addressing the trad~ offs ~nvolved in interfac~

design is presented. Section IV presents InterfaceGuidelines for future firing room software. Additionalsuggestions are included for a proposed windows version ofsuch software. section V contains concluding remarks.

1.2. The Role of the NTD

The NASA Test Director functions as thecoordinator of information during a launch countdown. Hereceives information from several diverse soUrces: firingroom clocks located on the wall in front of him and to hisleft, procedural information in 'hard copy from the OMI(Operations and Maintenance Instructions) and statusinformation received over the CIS (Operations InformationSystem). The current firing room clocks are universal time(UT), local time(local), countdown time (shuttle), windowremaining, post LOX drain back elapsed, APU hold timeremaining, time to T-O, and hold time remaining. Specificinformation about each launch, such as projected time oflift-off, launch window end,etc., is contained in a launchdocument for that particular mission. He also has access toclosed circuit television which' is directed at operationsoccurring around the launch pad~ One of the manyresponsibilities of the NTD is that of time management. Thatis, given that a hold has occurred in the countdown, The NTDmust decide upon the time to resume the count so that liftoff will not occur within a COLA or with little contingencytime. In making decisions concerning time management, theNTD has to carry out some arithmetic operations and base hisdecision on those calculations plus his knowledge ofspecific launch information. In addition, he uses knowledgeof approaches used in previous launches.

The LDSS consists of several parts: the timemanagement integrated display, the what-if capability, the

416

r................" 7(' situation assessment capability, and anomoly management.

The time management subsystem presents an integrateddisplay. That is, all firing room ,clocks are duplicated onthis display. Additionally, some clocks that performarithmetic are displayed. The launch window is shown alongwith any COLAS (collision avoidance). These representperiods in the launch window such that a lift-off duringthis period could results in a collision with anotherorbiting vehicle. contingency times are shown in the launchwindow as these also affect resumption from a hold. Thetime management display (shown in figure 1) calculates theadvisability of resumming from a hold and displays thisinformation in its resume window. A now bar displays thecurrent position in the countdown and a projected T-O bar isconsistently updated to reflect where lift Off will occur.The LOX DB is the contingency time currently deemed to bethe most constraining.

1.3 The Need for Human-Computer Interaction Knowledge

LOSS is designed to assist an NTO duringcountdown activities. Activities such as these areperformed on the order of every two weeks. Thisincludes simulations and actual launch activities. TheNTOs alternate with one NTO and one ANTO assigned toevery launch team. Therefore, these are casual users.The activity they perform, launching a manned shuttle,is a critical operation. The countdown activity is acognitively demanding activity that must be carried outin a real-time situation. This means that any toolsdesigned to assist the NTD and ANTD need to present theneeded information in such a fashion that it can beeasily comprehended and used. This project reports ona suggested redesign of the user interface for theLOSS. Accomplishing this redesign led to many tradeoffdecisions. This study also proposes a method forprioritizing guidelines in order to make consistentdecisions about trade-offs.

417

8

II. LDSS Revised Displays and Data Collection Results

2.1 TMID Revised Display

Figure 1 shows the original TMID screen andFigures 2,3 and 4 show the revised display. The largestchange was using one "window" on the display for informationconcer~ing the launch window. The original displaycontained a launch window which showed COLAS, a resumewindow which showed the advisability of resuming from ahold, and a LOX drain back window which showed the LOXcontingency time remaining. In addition, the displaycontained a "now bar" and a projected T-O bar, to show wherethe countdown currently was and where this meant that T-Owould occur. Interviews with members of the NTD staffindicated that there was confusion with using threedifferent windows, especially the launch window and theresume window. Therefore, the display was reworked usingonly one window - the launch window - and incorporating theinformation about COLAS, contingency time and advisabilityof resumming in this window and elsewhere on the display.The revised display uses the stop light coloring coding onthe launch window. Green indicates that the window is openat this point. Yelllow indicates contingency times and isseen prior to COLAS (indicated in red) and prior to the endof the launch window. The end of the launch window islabeled as such and nothing is displayed after this point.The "now bar" is coded to indicate the advisability ofresumming at this point in time. Green indicates thatresumption is safe (or that the countdown is not currentlyin a hold). Yellow indicates that a resumption at this timewould not have the full contingency time. Red indicatesthat resumming from a hold at this time would place T-O in aCOLA. There are two methods of finding out how long theperiod for resumming exists. One is by looking at thelaunch window to see the amount of green, yellow or red(also labeled) displayed below the projected T-O bar. Atthe beginning of the T-20 hold this information would haveto be obtained by scrolling the launch window.Additionally, this information is contained in a new clock"Time to No start". This will provide exact times thatindicate how long the current condition exists.

The time labels were also changed. The originaldisplay labeled all times, universal time and countdowntime. The revised display only labels significant times.That is, only the minutes in which a GLS event or a COLA or-i contingency time appears will be labeled with the univeraland countdown time. Countdown times are only labeled fortimes prior to T-O. Positive countdown times make little,sense for COLAS and contingency labels. The "now bar"always reflects the current time correct to the second.

418

Using a reducedare significantof the display.III.

amount of time labels will make those thatstand out and reduce the overall complexity

More will be said about this in section

9

The original TMID screen contained a scroll bar onthe left side of the display which was used to indicatewhich portion of the display was being viewed. The revisedTMID screen had more room so that the scroll bar could bemoved to the right hand side of the display in keeping withan OSF/Motif presentation. The scroll bar was designed sothat it resembles that of OSF/Motif. Recommended movementwith the display window should be via the scroll bar or byusing the up arrows and page keys. Either motion results inthe slider of the scroll bar being changed to relfect whatportion of the display one is viewing. In a directmanipulation version, the user should also be able toposition the slider to cause movement.

Color coding on the revised TMID screen isconsistent. Graphic information is color coded using thegreen (safe), yellow (cautionary), and red (warning) trafficlight metaphor. However, the graphic data has also beenlabeled so that color coding is redundant and therefore, canbe used by a visually impaired person. Text informationuses cyan for labels and place holder values. Whiteindicates information that needs to be input. Values ingreen are relevant; that is, data that has occurred. Yellowlabels and values depict information concerning a hold. Redvalues are values that indicate a warning situation exists.One color change that was made to the original TMID screenis that recommendation of using cyan as the color for datathat is not yet relevant. Previously non-relevant data hadbeen displayed in white. White should be reserved for veryimportant data. Data that is not yet relevant is notinvalid, but merely serves as a placeholder. Therefore,using the same color as labels (cyan) is recommended forsuch data. In general, no more than four colors should beused for alphanumeric information. The revised TMID screen(and other screens in LDSS) use cyan, green, yellow and red.As red is used very sparingly and only in the case ofwarning situations, the use of the three colors to displaytext, plus the background color of black, closely adheres tothis guideline.

The clocks in the original TMID screen wererearranged and "chunked" together according to function.For example, time in hold and hold time remaining werelocated together. This reduces the complexity for the userby presenting a few chunks of information rather than alarge number of information pieces. In addition, the clockswere arranged in order from top to bottom with those on thetop being the highest priority. This arrangement was basedon interviews with the NTD staff. This rearrangementallowed enough room on the display to include the maximumhold time remaining and the latest resume time. For timevalues that are not currently in use the recommendation is

419

10to use a default value of HHMM/SS or DDD:HHMM/SS as a promptto the user of the format of the clocks. This is due to thelack of room for labels on the clock values. A graphicdepiction of where the projected T-O falls in the launchwindow was also incorporated into the clock display. Thiswas purposely designed NOT to look like a scroll bar as itCANNOT be manipulated and is used to quickly show the NTOhow close to the end of the launch window, the projected T-Ois.

2.2 Finer grain TMID Screen

Feedback from the users indicated that as T-Oneared, the minute resolution on the TMIO screen was toocoarse to follow. A finer grain of time, namely a 10 secondresolution, is recommended. Figure 5 presents this newdisplay. The same format as the TMIO screen is used. Thetime bars in the launch window now become 10 second bars sothat one rectangular box per line is used. Time labels areindicated on whole minute entries. This display should beincluded in the menu and the user could select to view thisdisplay when a certain point in the countdown is reached.

2.3 Overview Display

Another new display that was developed during thisperiod of time is the overview display. The concept is thatthis display would present the entire launch window pictureto the NTO. The display, shown in Figure 6, could bederived from the mission parameters that are input prior tobringing up LOSS. The entire window is displayed withCOLAs, COBT, and built-in holds at T-20 and T-09. Thisdisplay could be used in a static fashion. That is, hardcopies of it could be printed and used by the NTO to get anoverall picture of the launch window. Figure 7 shows thatthe same display might also be used in a dynamic fashionwith the addition of a now bar and a projected T-O bar.

2.4 Test Parameters Display

Figures 8 and 9 present the original and revisedTest Parameters display. The original display listed thebuilt-in holds by time. Normally, searching would be doneaccording to the time of the hold. Therefore, these columnswere rearranged to facilitate that search. As LOSScurrently is designed to function from T-20, only thosebuilt-in holds at T-20 and T-9 need to be displayed. Theoriginal display contained only the start and end time ofthe launch window (unlabeled). The revised display containsthe start, end and length of the launch window, COLAS, andCOBT. In addition, the point of no recycle is now includedon this display rather than on the TMIO screen.

This same display could be used for input abouteach specific mission. The input display looks very much

420

11like the mission p~rameters display but contains informationon how to input new values. The data fields are displayedin white, indicating that these fields are to be filled inby the user. The fields indicate the number of positionsthat are to be filled in and the format of that data. Forexample, a length field in countdown time format would looklike HHMM/SS. A universal time format would be coded asDDD:HHMM/SS. In entering data, the user should not have toenter the fixed delimiters, e.g., : and I. These delimitersare shown in cyan to indicate that they do not have to betyped in. Movement through the display should automaticallyposition the user at the next number to be filled in,skipping over any fixed field delimiters.

An alternative suggestion is to retain values froma previous mission in these positions. This is feasible ifmany values do not change from launch to launch. Suchitems as LOX drain back and APU hold times and build-inhold times do not vary between launches. Therefore, sometime could be saved when entering these parameters. withthis method of entering data, the user should be able toretype only those positions within a field that need to bechanged. For example, changing 0004/00 to 0007/00 shouldnecessitate positioning the cursor on the 4 and retyping a7. Deciding on the method to use depends on the ability tosupport the correct interaction method and on the amount ofdata entry that the user could be spared. This methodshould be used only if the above interaction can beimplemented and if the users feel comfortable using cursorpositioning and retyping data. If this is not the case,then the first method should be used.

2.5 what If Display

The What If display allows the user to do somecalculation on various times that are adjustable. The endresult is to determine the maximum hold time remaining andthe latest possible resume time. These calculations arebased on COLA and contingency information in the launchwindow and the current countdown time and universal time. Apossibility that exists within these calculations isresuming the count and holding for a certain amount of timelater in the countdown. Figure 10 is the current version ofthe display. As the what if portion of the program is nowimplemented, three displays are used. Each display presentsa different default plan. One display reflects thesituation where no intermediate hold is used. The seconddisplay uses this intermediate hold situation. The thoughtbehind revising these displays was to create one generictemplate that would encompass both scenarios as well asother scenarios. Two possible revisions were developed.The first is a textual display and the other a graphic

/_." version.Figure 11 presents the textual display. The times

in the original display were presented in three columns,

421

12universa1 time, countdown time, and interval time., ,Thethought in revising the display was that the universal time~\and time intervals were most frequently uSed. Therefore, thecountdown column was eliminCited and the times were put intoa middle column. The labels were put into two columns". oneforevehts and the second for intervals. An up/dowh arrowbeside of selected items indicates that the user may changethe value of that item. The display allows for the user tochoose the countdown time of the intermediate hold. Thethought here is that by clicking on the selection ,arrows thetimes displayed would be the prescribed hold times, e.g., T-4M, T-2/55M, T-31S. The display includes start/end timesfor COLAS, but this information does not necessarily have tobe included. As with the original display the events areordered by time.

The graphical version (Figure 12) lends itself tofull direct manipUlation but could function in. the samemanner as the textual display thus allowing implementationto proceed in stages. In a full direct manipulation modethe user could position events by dragging themto.aposition on the time line., .The time value would be shown asthe item is being moved. This allows a fine~ 9raihofcontrol than the user has by dragging. The user could alsoclick on the up/down arrows and the item would be positionedon the time line relative to the value selected for it.Repositioning would only take place after the User haspressed "enter" to indicate he hai selected the desired /~~

value .. A first implementation ~nvolvingseiectionof valties I

in this manner is advocated before implementing the directmanipUlation interface. The benefit of this type ofinterface is that it allows the tiser to easily accomplishsuch tasks as "position T-O close to the end of the window".Figure 13 shows how the user could be notified ofconstraints. .

In addition, a suggestion is to allow the user tosetup several default plans, such as going to T-5M andholding for 5 minutes, prior to countdown and theri, selectingthose plans by selecting the label. Input of these,planscould be done via an input display that resembles the whatif display. Figure 14 presents a prototype.of an inputdisplay for default plans. These plans would be setup afterthe mission parameters have been input so that values forcontingency time and launch window end would be obtainedfrom there. These values cbtild thEm be changed in the •default plan. For example, the,NTD could construct. a defaUl.tplan that would necessitate asking for an extra five minuteson the end of the launch window. These plans couid then befiled and later retrieved and applied to the currentsituation.

Both the graphic and textual displays could bemade available to the user and he could choose to displaywhichever is more appropriate to his style of decisionmaking.

422

13

2.6 The situation Assessment Display

The situation assessment display containsinformation about the hold, the end of launch window,projected T-O and other information and makes arecommendation to resume or not based upon this information.This recommendation also includes the rationale for thedecision. This module is still being developed and hencethe information that will be displayed is not yet fUllyspecified. Recommendations for a display are shown inFigures 15 and 16. A template of information should bedeveloped and those labels should always be displayed~

Information that is not relevant to the present situationshould not be filled in. Furthermore, information should becolor coded so as to convey to the user the values in thetemplate that are contributing to the seriousness of thesituation. For example, if the projected T-O is within acola, that text would be displayed in red. The assessmentinformation would display information about resumming inred. This would allow the user to quickly assess that it isunadvisable to pick up by using the red indicators in theresume field. If he wishes to read the text to obtain moreinformation, he may do so. But he would be able to obtaininitial information via position and color. Further work isbeing done on this module. Using this initial display willhelp in defining a set of variables that should be examinedin any given situation.

2~7 Data Collection

The revisions that were made to the displays arebased upon guidelines and theories (Dumas, 1988; Galitz,1989; Gilmore, Gertman, and Blackman, 1989; Helander, 1988;NASA, 1980; smith and Mosier, 1986; TUllis, 1981) and upondata collected from the users. User data collection wasdifficult due 'to the workload on the NTD staff. Severallaunches were carried out during this research period andthe NTDs were engaged in those as well as the simulationruns prior to each launch. However, the following procedurewas carried out. The current system was used by an NTDduring a simulation, his verbalizations were recorded onaudio tape and notes were made about situations. Thefollowing items were noted during this process. Included isthe resolution of each item.

1. Qualitative information on window remainingshould be easily accessable.

Resolution: a graphical representation ofprojected T-O within the window was incorporated into theclock section.

2. COLA information is used for picking up at T-9M.

423

14Resolution: this information is shown in launch

window but consideration should be given to including theexact times on the labels.

3. The point of no recycle should be included onthe test parameter screen but does not need to be includedwith clock information.

Resolution: this has been incorporated in therevised TMIO and test paramater display.

4. The major benefit of the what if display wasenvisioned to be taking the launch window end and workingbackwards from this. The time between launch window end andprojected T-O was seen to be dependent upon contingency timebut not entirely.

Resolution: The textual version of the revisedwhat if display breaks this time interval into two parts:contingency time plus an additional time. Both should becapable of being changed to give a resulting interval.

5. The clocks that are used differ in importancedepending on the countdown time of the hold. At T-20M, thehold time remaining is most important. At T-9M the windowremaining is important. At T-5M and under the hold timeremaining (which increments at this point) is important.

Resolution: consideration should be given toincorporating code to highlight the appropriate clocks atthese times in order to direct the user's attention.

6. When scrolling through the launch window itwas easy for the user to loose his orientation.

Resolution: moving the scroll bar to the righthand side of the display and making it more visible isimportant. The user can use this information to jUdge whichportion of the display he is viewing. This has beenincorporated into the revised TMID screen.

7. A method of intially putting in parameters foreach launch and for changing them during countdown should beprovided.

Resolution: A display has been prototyped forthis purpose." This feature is currently on the list ofitems to incorporate into LOSS.

The second method of data collection involved aninitial series of displays, namely the TMIO screen, whichwere revised based upon informal comments and guidelines.Several situations were setup in order to obtain informationabout interpretations of specific decision making processes.These were shown to an NTD and discussed. Based upon thisinformation, further revisions were made and again shown tothe same NTD. This set of displays was distributed to othermembers of the NTD staff along with a questionnaire. Thisdata yielded the following information:

1. The2. The

has been changed.into LOSS.

clOcks provided are sufficient.LOX drain back clock in the firing roomThis change will soon be incorporated

424

153. Make sure that users understand which values

(,-, in the clock section are static (Launch Window End) andwhich are computed. This needs further discussion with moremembers of the NTD staff. A color code could possibly beused here or the positioning could be changed.

4. Reduced time marks are suficient.5. A finer grain of time (second resolution) was

suggested. This is provided in two ways. First, when theLDSS runs the time shown is dynamic and updated on a secondby second basis. Secondly, a new display has been developedwhich the user could switch to close to lift off whichprovides a ten second resolution.

6. The color coding of the graphic display wasdeemed useful and clear.

7. Color coding for the data values and time wasquestioned. The explanation of the color coding should beseparated when presented to the user. For graphics, greenrepresents safe, yellow cautionary and red represents awarning. For data values, cyan represents not yet relevant,green represents a valid, relevant data item, yellowrepresents a hold condition. Therefore, all clock valuesshould appear either as cyan or green. The only exceptionis time to T-O which should appear in yellow when there is ahold in the countdown. On the timeline of the TMID screen,the countdown times will be displayed in yellow when thereis a hold. UT, CDT and event labels that have happenedshou+d change from cyan to green.

8. GLS milestones should be labeled to refecttimes accurate to seconds. This has been incorporated onthe revised TMID screen.

9. Displaying latest resume time on TMID screenis useful. This was used in the decision making process.Therefore having this time available on the TMID screen willreduce movement between displays.

10. A suggestion was made that it would be usefulto have more information on time crital actions such asstart recorders, APU Fuel ISO's, etc. This information willbe displayed ~n a planned configuration management display.

2.8 complexity Issues

In predicting a user's ability to easily obtaininformation from a display several items must be taken intoaccount (Tullis, 1981). The number of items in a group andthe average size of the group is one determining factor inlowering search times. Results by Tullis suggest that theoptimum range for alphanumeric displays is 40 groupsaveraging 4.9 degrees in size. Local density, how tightlypacked a screen is, and layout complexity (the alignment) ofthe display also affects performance. Also important, thesetwo issues affect sUbjective ratings by users. These issueswere examined in the displays that are suggested forrevision. The grouping of items and the size of the groupswas considered somewhat in the revisions. This was mostly

425

16in the clock information. Rather than keeping each clock asa separate entity, information was grouped according tofunctionality. The size of the groups was determined by theinformation presented and therefore, other rearrangementswere limited.

Complexity was analyze4 for an original TMID and arevised TMID screen. As the information presented on thesedisplays changes, various c,omplexity measures will beobtained. For this analysis, two displays were chosen thatrepresent one of the more complex displCiYS. Using a easiercalculation than the original Tullis calculation frominformation theory (Galitz, 1989) the following are summed:the number of data fields, the number of horizontalalignments (columns) and the number of vertical alignments(rows). For the original TMIDscfeen shown in Figure 17this produces a complexity score of 177. For the revisedTMID screen, the complexity score is 153. This revision isa 13.6% reduction from the original.

Density of a display is the proportion ofcharacters displayed on the screen as opposed to the amountof blank screen. For density calCUlations, the screen wasconsidered without its border. Dimensions were 70 by 28reSUlting in 1960 positions. F.or the original TMIDscreen,the density was calculate<:ito be 5.2 percent. The densitycalculation for the revised screen wa.s 40 percent.Although, this revision is still higher than the overallrecommended level of 25 to 30 percent the fact that aportion of it is graphical should still result in a usabledisplay. Further reduction is impossible as a principleadhered to in redesign was to reduce the amount of movementbetween displays. Therefore, this trade-off of higherdensity versus time to switch displays was made.

426

17III. Trade-offs in organizing Human-Computer Interaction

Knowledge

3.1. Introduction

currently, a large body of human-computerinteraction knowledge exists. Most of this exists in theform of recommendations and guidelines. One of thedifficulties in interface design is transferring thisinformation to interface designers in such a manner that itcan be incorporated into the design. Scapin (19~1) examinedthe problems encountered when trying to implement human­computer interaction knowledge into the design. Scapindeveloped a framework for organizing this knowledge. Thissection discusses this framework, the problems that arise inusing this framework, and a method for resolving one of theproblems, namely, that of weighing trade-offs.

3.2 organization of HCI Data

Scapin collected Hcr data with the intent oforganizing it into a database accessible by attributes.Attributes that were used for characterizing Hcrrecommendations were: criterion or criteria describing therationale, the level (conceptual, semantic, syntactic,lexical) that describes the outcome and rationale, and thetype of interface object to which the recommendationapplies. Several conclusions were drawn by Scapin duringthis process. He found that a recommendation in theliterature could lead to more than one recommendation interms of implementation. Many recommentations were toovague and needed elaboration in order to be useful inimplmentation. Additionally, different authors useddifferent wording for the same recommendations necessitatingtranslation to a stable vocabulary.

Scapin proposed a generic deciphering framework ofthe form:

if (a premise) ... (using criteria) ... then (a Conclusion)

The criteria used are:Compatability with past habits and skills of the

user, between input and output, between noncomputerizedsupport and software, and with interface standards.

427

18consistency in location, format, syntax and naming

conventions.User Workload with respect to minimizing mental

and physical effort required.Adaptability or the capability of the interface to

adapt to various user actions.User Explicit Control allows the user to control

the dialogue and to explicitly enter information.Significance of codes or use of labels, codes, and

abbreviations that are meaningful to the user.Guidance so that the user can identify what he can

do next. This includes feedback and clarity in display.Error Managment is the attempt to prevent or avoid

errors and to give meaningful feedback when errors do occur.

Conclusions can be of two types: design activity whichrepresents a conclusion or specific activity or an actionitem which is the type of activity required to apply a rule.

scapin notes intrinsic and usage problems involvedin applying human factors knowledge. Intrinsic problemsconsist of recommendations that are incomplete,recommendations that are too general to be useful andrecommendations that lack robustness (apply only in limitedcontexts). Usage problems are accessing recommendations,making trade-offs in deciding which recommendations toapply, and the varying degree of detail of therecommendations.

3.3 A Method for prioritizing Trade-offs

Recommendations do not come organized in ahierarchy. Therefore, the designer will often be faced withseveral approaches, each giving priority to differentrecommendations and resulting in different solutions. Inorder to balance trade-off decisions, several issues arise.The first is what factors should be used to determineweights for guidelines. The second issue is how to ensurethat these trade-off decisions are made consistently. Themethod presented here builds on Scapin's work. As discussedin section 3.2, scapin presented eight criteria andsuggested the types of guidelines that pertain to each. Inorder to prioritize decision making, these criteria are usedand weighted according to some known user and taskinformation. During the evaluation of LOSS, the followinginformation was used to prioritize guidelines: type ofuser, attention time, timing of task, critical aspect oftask, and critical aspect of software. Each of thesecategories can have several different values which woulddictate prioritizing different criteria and hence,guidelines. The following discussion is a suggested set ofvalues and emphasized criteria.

Type of user refers to the novice, casual, expertclassification. For a novice user the primary criteria

428

,~, '\

I

19would be guidance. Error management, compatability withpast habits, significance in codes, consistency workload andexplicit control would be secondary criteria. A casual userwould dictate the same criteria but consistency and workloadwould become primary criteria along with guidance, errormanagement, and compatability. Expert users would be moreconcerned with adaptability than any of the other criteria.

The attention time of the user is also animportant factor in determining trade-offs. What elsecompetes for the user's attention while dealing with thesoftware? Suggested values range from undivided attentionto moderate attention to little attention. In the case oflittle attention, where the user is attending to many othertasks and using the software primarily for support, issuesof consistency, guidance, workload and significance of codebecome the primary criteria. Compatability becomes asecondary concern. In the case of moderate attention thesame criteria are important but could be consideredsecondary concerns rather than primary concerns. If theuser has undivided attention to devote to the software thenadaptability could be classified as a primary concern. Theuser has time to discover various aspects of the system andis able to choose a method of performing a task that is moresuited to his individual style.

Task considerations should include whether thetask is a real-time or nonreal-time task. That is, is thesoftware supporting a shuttle launch or is the software anincome tax spreadsheet? This consideration is similar tothe attention of the user consideration in that the primarycriteria for real-time tasks should be the same as forlimited attention: workload, guidance and consistency.Nonreal-time tasks place adaptability as a primary concern.The critical issue of the task involves the importance ofthe software to the task. That is, can the task becompleted without use of the software? critical softwarewould dictate that error management be a prime concern withguidance as a secondary concern. A second concern is thenature of the task. Software that aids in landing aircraftwould be deemed more critical than software for producingslides. criteria important for this type of criticalsoftware would be: significance of codes, guidance, andconsistency. The emphasis is placed on the ability todorrectly interpret information presented.

The method for prioritizing criteria and hence,guidelines, involves collecting this information concerningusers and tasks. Then individual criterion is given twopoints for each time it appears as a primary concern and onepoint each time it appears as a secondary concern. Thecriteria with the highest scores should then be used as thedeciding factor in trade-off decisions. Guidelines thatsupport those high scoring criteria should be utilized overguidelines supporting criteria receiving lower scores.

The method used here needs further refinement toensure that it correctly addresses all values for suggested

429

20categories. Nonetheless, this method was used quitesuccessfully in making consistent trade-off decisions duringrevision of the LDSS interface. In this case, the userswere casual, the attention of the users limited, the taskwas real-time, and critical but the software is not criticalfor success of the task. Using this scheme an interface wasproduced in which priority was given to guidelines whichenhance the users' ability todiseern information from thescreen and to minimize the amount of interaction required bythe user. consistency and guidance were the highest rankingcriteria followed by work load and significance of codes.Little effort was put into making the system flexible. Asthe amount of interaction was limited, little effort was putinto error management. In order to further refine thismethod and to determine its robustness, it will be necessaryto evaluate more interfaces which have different tasks andusers. A successful method would be capable of producingusable interfaces for a wide range of user and taskcharacteristics.

430

21IV. CONCLUSIONS

Suggested revisions for the LOSS interface havebeen developed using a combination of empirical informationand application of human-computer interaction knowledge.The process of developing these revisions was tracked andled to a method for prioritizing criteria for use in trade­off decisions. As LDSS will now be field tested in the KSCfiring room, the actual usability of the interface can beassessed. Due to the nature of the ANTD/NTD job,information concerning the use of the interface will have tobe collected in a nonintrusive fashion. As the use of LDSSis optional and will require a change in the ANTD/NTDprocedures during countdown, the initial and sustained useof LOSS will be a major indication of the usability of thesystem. Additionally, information must be collected afteruse of the system. This information should discriminatedbetween missing information, misinterpreted information andmistrusted information. Audio tapes of countdownactivities, especially S0044 simulations, along withobservations by the LOSS staff can be used to obtain thisinformation. In addition, post countdown interviews withthe ANTD and NTD are very valuable. Any changes that aremade to LDSS as a result of these observations andinterviews should be documented in the form of guidelines.New software development will benefit in the form of timesavings and monetary savings from use of these guidelines.Utimately, the users of these software tools will benefitfrom the increased usability of newly developed tools.

431

22REFERENCES

Dumas,J.,Designing User Interfaces for Software, PrenticeHall, 19;88.

Helander,M., Handbook of Human-oomputer Interaction, NorthHolland: The Netherlands, 1990.

Galitz, W., Handbook of Screen Format Design, QEDInformation Sciences, Inc. Wellesl:ey, MA, 1989.

Gilmore, W., Gartman,D., and Blackman, H., user-ComputerInterface in Process Control, Academiicpress, 198~.

Scapin, D., Organizing Human Factors Knowledge,International Journal of Human-Computer Interaction, Vol. 2,No.3, 203-229.

Smith, S. and Mosier, J.,Guidelines for Designing UserInterface Software, ESD-TR-86-278.

Spacelab Display Design and Command Usage Guidelines., NASA,MSFC-Proc-711A 1 April, 1990.

Tullis, T.,The formatting of alphanumeric displays: Areview and analysis. Human Facotrs, 25, 1983, 657-682.

432

UHIUERSAL TIME COT TIME T0 T-O005:1020/00 -00:0009/00 0013/00

HliJLOIHG1018/00 I!lI!lSllIGlI!l -OS/OOH HliJLO TIME REM1019/00 I!lSStDl!l1!l -OS/OOH 0004/00

01010101010

1021/00mmmmml!l -OS/OOH TIME IN HIDLO1022100 I!ltDl!ll!lGlI!l -OS/OOH 0006/001023/00 I!lmtDtDtDl!l -OS/OOH1024/00 IIIIII -09100 GlS SEQ UIH00U REM1025/00 ..1111 -08/00 IDAA RETRACT 0111/001026/00 IIIIII -07/001027/00 IIIIII -06100 lIDX ORAIH BACK1028/00 IIIIII 1lI1l11!l1lltDllI -05/00 APU START 0004/001029/00 111111 llImmriHiHll -04/00 PURGE SEQ 41030/00 .11111 I!ll!ll!ll!lllll!l -03/00 lmx PRESS PRliJJECTEO T-O1031/00 1..111 Elmmmma -02100 LH2 REPlEHISH 005:1033/001032100 111111 G'G'l!lmmG' -01/00 GLS AUTID SEC

..11..- III I I • I I II I ~ AUT@ SCR@LL,8H1034/00 1IIIII II1II1 1II11I +01/00 SELECT MISSIIDH1035/00 1111.1 1I1II1 1IIII1 +02100 RESTART1036/00 111••1 1•••11 1.1111 +03/00 EXIT1037/00 11•••1 1••1.1 +04/001038/00 1••111 11.111 +05/00 PT HliJ RECYCLE1039/00 11.111 I11II1 +06/00 005:0932/001040/00 .11111 1IIIII +07/001041/00 II1III 111.11 +08/00 OEMliJ1042/00 IIIIII IIIIII +OS/OO OEMliJ

Figure 1: original TMID

MAXHIDLD TIME0015/00

LATEST RESUME005:1043/00

TMWI ~ TM5A ~

LOST ELAPSEDHHMM/S5

CliJHTIHGEHCY TItlEHHMM/SS

: HIDLO TIME REM0001/00

: TIME IH HliJLO0009/00

TIME TID H@ START0002100

H@LOIHGCOT

-00:0009/00

1029/00 -09/00 GLS SEQ1030/00 -08/00 0AA RETRACT 7/30

1033/00 -05/00 APU START1034/00 -04100 PURGE 5E041035/00 -03100 L@X PRESS 21'551036/00 -02100 LH2 REPLEH 1/571037100 -01/00 GLS AlITID SEQ 0/31

....111040/00 •••••• CIDNTINGENCV ­......

111111111111

1044/00 IIIIIICIDLA- .111.11046/00 ..1111- 1111111048/00 I •••• IC@NTIHGEHCV ­

..1111

..111111111.

1052/00 LW END

1038/00 .111.1 . --00/00 ~PR0JE(TEO T-O

1028/00 MffiU £AR ., '., ;: --(I'j/OO - - ,-,. " ' .-

LIT005:1028/00

TIME Tffi T-O0010/00

1------------+----------,-; PRIDJECTEO T-0005:1038/00

I 01---1UIHO@U REtl

0024/00. LAUNCH UIH END

005 :1052100

Figure 2: Revised TMID

433

UT COT TIME TID T-O005:102S/00 -00:0020100 H.IDLOIHG 0031 roo

PRIDJECTED T-O,005:1058/00

- -- - I 0 I- - WIHDIDW REt!- - 0130/CIO- - LAUHCH WIH END- - 005:1159/00- -- - HIDLD TIME REM- - 0001/00- - TIME IN HIDLD1040/00 -09/00H 0011/00- -OS/OOH TIME TID HID START- -OS/OOH 0005/00- -09/00H- -09/00H LOBT ELAPSED- -09/00H HHMM/SS- -09/00H CIDNTINGENCV TItlE- -09/00H HHMM/S5- -09/00H- -OS/OOH MAX HIDLO TIME1050/00 -09/00 GtS S·EQ 0015/001051/00 -08/00 ~AA RETRACT .7155 LATEST RESUME- - 005 :104 31001053/00 - • TMUI ~ TMSA ~

24

Figure 3: Reviged TMIO, T-20M

UT COT TIME TID T-O005:1029/00 -00:0020100 HUDING 0.031/00

PRlDJECTED T-O,005:1058/00

- -- - I 0 I- - WIHDlDW REt!- - 0130/00- - LAUHCH WIN END- - 005:1159/00- -- - .. H.IDLD TIME REM- - 0001/00- - TIME IN HlDLO1040/00 -OS/OOH 0011/00- -09/00H : TIM'E TID ·H.ID 5TAR T- -09IDOH 0002100- -OS/COH- -OS/OOH LDBT ELAPSED- -09/00H HHMM/SS- -09/00H ClDNTINGEHCV TIME- -OS/OOH HHMM/SS- -09/00H- -09/00H MAX HfilLOTIME1050/00 -09100 GLS SEQ 0100/001051/00 -08/00 filAA RETRACT 7/55 LATEST RESUME- - 005:1129/001053/00 - • IMUI I§I TMSA ~

Figure 4: Revised TMID, T-20M, Unable to Resume

434

LOST ELAPSEDHHMM/SS

CIDHTIHGEHCV TIMEHHMM/SS

HIDLD TIME REMHHMM/SS

lIME IH HIDLOHHMM/SS

TIME TID HID STARTHHMM/SS

MAX HIDLD TIMEHHMM/S5

LATEST RESUMEDDD:HHMM/SS

• TMWI ~ TMSA ~

IIII•1039/00 IIIII•1040/00 IIIIII

1041/00 • C0HTINGENCVII

1038/00 .' -00100 PR@JECTEO T-O

1037/29 H0U BAR -00131

UT COT TIME TID T-O005:1037/29 -00:0000/31 CmUHTIHG 0000/31a-::-::--:..-------+---------r PR0JEC TED T-O

005:1038/00

...------1..,WIHOIDWREM

0016/00LW EHD

005:1054/00

Figure 5: TMID with fine grain resolution

.~-"

(

LAUHCH WIHDIDW IDUERUIEW 5TS-43

HR/MIH0 1 2 3 4 50123456789 0123456789 0123456789 0123456789 0123456789 0123456789

14 11111111 Ily 1111111 0]]1111111 IllY 111111t tT-20H T-20 T-09H T-09

15 m.......1 IIU...III .......... ••1...1••• ••1111.... ..........t

LW lilPEH16 1111111111 1111111... 11111111.1 1111111111 111111.111 1111111111

17 1111111111 1111111111 IIIII.UII 1111111111 1111111111 1111111111tCIDST

18 1.11111111 1111111111 1111111111 I1I1111111 111111.111 ••1••••••1

19 IIIIIIIXXX XXXXXXXXXX XXXXXXXXXX XXXXXXXXXX XXXXXXXXXX XXXXXXXXXXt

LW CUl5ELAUNCH IHHOIDW START SHlP LEH APU HlilLO

lilPEH 002: 1502100 ClilLA1 OOO:HHMM/SS DDD:HHMM/SS HHMM/SS 0007/00CLIDSE 002:1906/00 ClilLA2 OOO:HHMM/SS DDO:HHMM/SS HHMMJSS LOBTLEHGTH 0404/00 cmBT 002:1317100 002:1732/00 0415/00 0004/00

Figure 6: overview Display

435

LAUNCH WINDIDW IDUERUIEW STS-~3

26

HR/MIHo 1 .2 . 3 .. ~ .. 5 ...o123~56 789 0123.4 56 7890123~5~.·?181910 123456789.101234567890123456789

14 11111111111 0IlJil'.•.• 'II.". . •1I11111~.... • ......¥ID. 111 111'.....··mT-20H T-20 T-09H . .T-09 LW IPEH151......... • ,.U U .

Cl1lLAl16 •••••••••••••••••••••••••••••1 1•••••",. 1,.1.,••••••••••••••

17 ••1.,.,••1 .11111111•••••11••••••1••••••1 •••11111•• I•••••••••t tCl1lLA2 C@BT

18 1111111••1 111••••111 1111111111 11111••111 .111111'1•• 1.1•••11.1

19 111.111111 IIXXXXXXXX XXXXXXXXXX XXXXXXXXXX XXXXXXXXXX XXXXXXXXXXtLW CLIDSE

LAUHCH WINDIDW START STIDP LEN APU HIDLDIDPEN 024:1455/00 CIDLAI 024:1540/00 024:1543/00 0003/00 0007/00CLIDst 02~:1912JOO CIDLA2 024:1709/00 02~:1711/00 0002/00 LDBTLENGTH 0417/00 CIDST 024:1309/00 02~:1724/00 0415/00 0004/00

Figure 7: Dynamic Overview Display

TEST PARAMETERSDE MID

L0CATI0H: PAD I MlP IUEHICLE: lDU-I04ATLANTIS

BUILT-IN HIDLO TIMES

0800/00 SIH @ T-27H0800/00 BIH @ T-f9H1328/00 BIH @ T-IIH0100/00 BIH @ T-6H0200/00 BIH @ T-3H0000/00 BIH @ T-IH20M0010/00 BIH @T-20M0010/00 BIH @ T-9M

CIDLA 0005:1048/00 005:1051/00

TCIO: SB037Al

INITIAL APUHlilLO TIME 0007/00

INITIAL LIDX DRAIHBACK 0001;/00

INITIAL U.lX START SIDX 0005/00

INITIAL COT -00: 0021100

INITIAL PR@JECTED T-O 005: 1033/00

LAUNCH WINDl1lW 0095:1033/00 095:1131/00

Figure 8: original Test Parameters Screen

436

MI5SUN 5TS48

UEHICLE: lllU-XXX LlllCATIlllN: PAD X MLP X TeID: XXXXXXX

BIH LENGTH APU HIDLD TIME 0007/00LOBT 0004/00

T-20M 0010/00 LlllX START BlllX 0000/00T-9M 0010/00 PR0JECTED T-O 000:0000/00

PT N0 RECVCLE 000:0000/00

START END LENGTHLAUNCH WINDIDW 000:0000/00 000:0000/00 0000/00

CIDLA 1 000:0000/00 000:0000/00 0000/00C0LA2 000:0000/00 000:0000/00 0000/00C0LA3 000:0000/00 000:0000/00 0000/00

CIDBT 000:0000/00 000:0000/00 0000/00

Figure 9: Revised Test Parameters Screen

UNIUERSAL TIME COT TIME TID T-O005:1017/00 -00:0009/00 0016/00

H@LDIHGH@LD TIME REM

0007/00

MAX H@LO TIME REM 0056/00 TIME IN H0LDLATEST RESUME TIME 1113/00 0003/00

IHH001.J REMI--COT DESCRIPTUH UTe INTERUAL~ 0114/00

C0NTING TIME0004/00

-09/00 H0LO START TIME 1014/00TIME IN HlllLD 0003/00 PR@JECTED T-ON0W 1017100 005: 1033/00MAX HlllLD TIME REM 0056/00LATEST RESUME TIME 1113/00

o -05/00 H0LD START TIME 1117/00TIME IN HIDLD o 0005/00RESUME TIME 1122100

PT NID RECYCLE-00/00 DESIRED GMT @F T-O 1127/00 005: 0932100

C0HTINGENCY TIME o 0004/00LAUNCH UINO@W END 01131100

DEM@

Figure 10: Original What If Display

437

UT005:1028/00

EUEHTS

28

HIDU 0000/00DODO/DO

LATEST RESUME TIflE HHMM/5S

HHMM/SS

IHTERMEDIATE HIDLO @ HHMM/SST-OO ~ HHMM/SSPR0JECTEO T-O

CIDLAl

C@LA2

UI END

PLAN A I§]

~ HHMM/SS

HHMM/SSHHMM/SS

HHMM/SSHHMM/55

~ HHMM/5S~'HHMM/5S

PLAH I§]

TIME IHHIDLD

MAX HIDLD TInE REM

HIDLD TIMEHIDLD TIME REM

0001/00TIME IN HIDLD

0009/00TIME T0 N0 START

0002100

LDBT ELAPSEDHHMM/SS

CIDHTIHGEHCV TIMEHHMM/SS

Figure 11: Revised What If Display, Textual Version

UT CDT000:0000/00 000:0000/00 STATUS

WHAT IF TIMELltiE 0000/00CilLA CIDBT 1;1. ;1. ;1.

,,,'

t

11. ~RMCTEO r tHIDLO START LW EHD... 0000/00 ...TIME 000100 0000/00 T-O 0000/00 T T

RESUME LATEST RESUME0000/00 ~ TIME DODO/DO ~ C@HTIHGEtiCY

1 1 DODO/DO ......TIME IN HIDLD

t0000/00 ~ !MAX HIDLDTIME REM0000/00 ......

DEfAULT PLAHS

PLAH A ~ PLAH C ~

PLAH B §J PLAN 0 I§]

Figure 12: Revised What If Display, Graphical Version

UT000:0000/00

COT000:0000/00 STATUS

WHAT IF TIMELIHEcmaT...

0000/00

1

IiLW EHZ~ 0000/00 ~

CIDHTIHGEHCYDODO/DO ~

ITIME IHDODO/DO

tHIDLO STARTTIME 000/00 11 PR8JEtTEO

0000/00 T-O 0000/00

RESUME LATEST RESUME0000/00 * TIME 0000/00 ~1 1

HIDLO~ ! t

MAX HIDLOTIME REM0000/00 ~

DEFAULT PLAHS

PLAH A @]

PLAH B lID

PLAH C @]

PLAH 0 lID

Figure 13: Revised What If Display, Graphical showing Constraints

UT COT TIME TID T-O005:1028/00 -00:0009/00 H@LDIHG 0010/00

PRIDJECTED T-OSELECT UALUES FIDR OEFAULT PLAHS. THEH SELECT LABEL 005:1038/00F@R PLAH.

I 0 IWIHO@W REM

0024/00EUEHTS IHTERUALS LW EHO

005: 1052100

IHTERMEDIATE H0LD @ HHMM/SS HffiLD TIME REMFOO ..

~ HHMM/SS HIDLD TIME 0001/00...TIME IH H@lD

0009/00~ HHMM/SS CIDHTIHGEHCY TIME TID H@ START

0002100

LW EHO ~ HHMM/SS LOST ELAPSEDHHMM/SS

APU H6lLO TIMEHHMM/SS

MAX HIDLO TIME0015/00

LATEST RESUME@l ~

005:1043/00PLAN A PLAH PLAH C @l TMWI @l TMSA @l

Figure 14: What If Default Input

439

30

HUD TIME REMHHMM/SS

TIME IH HIDLOHHMM/SS

TIME TID HID STARTHHMM/SS

T-9MEXTENDEDLIDNGHIDHEWITHIH CIDHTIHGEHCV~f A CSLA

ASSESSMENT

TIME:HIDLD:LAUHCH WIHDilW:CffiHTINGEHCV: .PRIDJECTED T-0:

UT COT TIME TID T-O005:1035/15 -05:1009/00 HIDLDING HHMM/SSI-----------------------~~---.---.--- PRIDJ ECTE 0 T-0

SITUATIlDH ODO:HHMM/SS

I • IWIHDIDW REM

HHMM/SSLAUNCH WIN SUP

OOD:HHMM/SS

RESUME: UHADVISABLE

WHY: WILL NIDT HAVE fULL ClDNTIHGENCY TIME PRIIDRTID C@LA

LlDX DRAIN BACKHHMM/SS

CIDNTIHGEHC\, TIMEHHMM/SS

MAX HIDLD TIMEHHMM/SS

LATEST RESUMEODO:HHMM/SS

TMWI @ TMSA @

Figure 15: Revised Situation Assessment

HlDLD TIME REM0001100

TIME IN HIDLD0009/00

TIME TID HID START0002100

COT-00:0009/00

HlD CilNSTRAINTS UNTIL 0000/00

T-9MEXTERNALLIDNGLIDNG

TIME:H@LD:LAUHCH WIHO@!.l:MAX HIDLD TIME:CIDNTIHGEHCY:PRIDJECTED T-O:

SITUATHlN

UT005:1028/00

TIME TID T-OH@LDIHG 0010/00

.......--_-'-- ....L- ~ I PRIDJEC TED T-O005:1038/00

I 0 IWIHDIDW REM

002Lt/OOLAUNCH WIN END

005:1 0521 00

ASSESSMEHT

RESUME: ADUISABLE UHTIL 0000100

LDBT ELAPSED0000/00

ClDNTIHGEHCY TIME0000/00

WHY: HID CIDHSTRAIHTS MAX HIDLD TIME0015/00

LATEST RESUME005:1043/00

T"UI @ TMSA @

Figure 16: Revised situation Assessment, Display 2

440