Embed Size (px)
Citation preview
Massachusetts Institute of TechnologySloan School of Management
Executive ProgramsProf. Tom Allen, MIT
Knowledge Management
• Gaining Knowledge
– Technology Transfer • Between Organizations
• Within Organizations
– Gatekeepers
• Disseminating Knowledge– Technical Communication
• Organization Structure
• Physical Structure of Facilities
What do we know about technology transfer?
• It is a 'people process'.
• Transferring documentation is, at best, an auxiliary process.
• People must be in direct contact and understand each other to transfer knowledge.
• The best 'package' for knowledge is the human mind.
• Moving people is the most effective way to move knowledge
• This can imply either organizational or geographical movement.
• Organizational boundaries impose a serious barrier to the transfer of technology
• This is due to the development of different organizational cultures.
The Context of the Study
X X
X X
X
Project Team
Organization
'Twin' Projects
X X
XX
X
X X
XX
X
Company 'B'Company 'A'
Sources of Technology
X X
X X
X
Internal Staff
Outside Experts
Literature
Literature & Documentation
X X
XX
X
InternalStaff
OutsideExperts
Literature18%
4%
People as Sources of
Technology
X X
XX
X
InternalStaff
OutsideExperts
Literature
2X
1X
People Outside
People Inside
All Written Material
0
Proportion of Messages
Technology Sources for Product Development Projects
10 20 30 40 50 60
People as Sources of Technology
OutsideExperts
Positive
Negative X XX
X X
InternalStaff
Literature
Customer Evaluation of Solutions as a Function of Idea Source
High Low
Internal
External
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Pro
po
rtio
n
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Pro
po
rtio
n
Evaluation of Solution Source of Idea
'Boundary Impedance' of the Organization
OutsideExperts
Negative
X XX
X X
InternalStaff
Literature
Science and Technology
• Science is Universal.
• Technology is Local.
Technology
• Technology is defined in terms of:
• The Business Goals
• The Marketing Strategy
• and most importantly,
• The Culture
• of the organization in which it is developed.
• Technical problems are thus defined in terms of that culture and its system of values.
The Local Nature of Technology
• This implies that:
• Anyone outside of the organization cannot fully understand the way that those within the organization define technical problems without understanding the organization's culture.
• This difficulty in understanding the problem is the principal barrier to technology fransfer.
• Barriers of this sort arise any time that we try to transfer knowledge across organizational boundaries.
• It thus holds true for internal communication as well as communication with other organizations.
• It is one of the causes of poor interfunctional relations in organizations.
Performance in Transferring Designs to Manufacturing
as a Function of CAD System Use for Communication
0 10 .2 0 .4 0 .6 0 .8 1 . 2 1 .4
L o w
H ig h
Q u
ali
ty o
f T
ran
sfe
r (F
ew
er
E C
s)
Mean Com m unications per Day
(T hrough C AD System )
p < 0.02
Using a Common Reference to Reduce Ambiguity in Communication
Product Development Manufacturing
Engineering
Effectiveness of Strategies for Reaching Common Understanding of Problems by Product Development and
Manufacturing Engineering
1
3
5
7
Mgt I
nterv
entio
n
Freque
nt Mee
tings
Sho
w &
Tel
l
Tests
Co-
Loca
tion
Pe
r f o
r m a n
c e S
c o
r e
Manufacturing Engineering
Product Development
Prototype
Partial Layout of the BMW Forschung und Ingenieurung Zentrum
The Effect of Transfers
A B
N1
Continuing Relations
A B
N1
Potential Contacts
More Continuing Relations
A B
N1 N2
Potential Contacts
Referrals
A B
N1 N2
Referrals
A Typical Technical Communication Network
(N > 600)
Netgraph of Communication Among Software Developers
20
Year
Olds
30
Year
Olds
40
Year
Olds
50
Year
Olds
Netgraph of Communication Related to Age
20 30 40 50
Year Year Year Year
Olds Olds Olds Olds
to Organizational Structure
C3
A1 A2
B1
C2 C1
C4
C5
C6
C7
C8
C9
D1
D2
A3
A1B1C1C2C3C4C5C6C7C8C9D1D2
De
pt
CD
ep
t D
Netgraph of Communication Related
A2 A3
De
pt
AD
ep
t B
Physical Location
Fl 2
Fl 1
Fl 4
Fl 2
Fl 3
Fl 1 Fl 2 Fl 1
Fl 1
Fl 2
Fl 1
Fl 6
Fl 7
Fl 5
Bldg. A1
Bldg. B1
Bldg. C3
Bldg. C2
Bldg. A1
Bldg. B1 Bldg. C1
Netgraph of Communication Related to
Bldg. C1 Bldg. C2 Bldg. C3
Communication
Related to
Organizational
Structure
Netgraph of
Manufacturing Plant
Manufa
ctu
ring
Pla
nt
Communication Mapping
• location Order by
Communication Mapping
•
Project
Development
Process
Order by
Among Groups in Laboratory ‘K’.
Gro
up
s
Netgraph Showing the Low Level of Communication
Groups
Communication Network in a Small Laboratory
of Refereed Journals
Laboratory 'A'
Laboratory 'G'
Laboratory 'E'
Laboratory 'L'
Laboratory 'M'
Laboratory 'H'
Laboratory 'S'
Laboratory 'T'
Laboratory 'U'
Laboratory 'V'
0 5 10 15 20
Mean Journal Readership
High Communicators
All Other Staff
p <
High Communicators Compared with Colleagues in Readership
p < 0.001
p < 0.02
p < 0.01
p < 0.05
p < 0.05
p < 0.01
p < 0.01
p < 0.01
0.01
N.S.
High Communicators Compared with Colleagues in Terms of
Laboratory 'A'
Laboratory 'G'
Laboratory 'E'
Laboratory 'L'
Laboratory 'M'
Laboratory 'H'
Laboratory 'F'
0 1 2 3 4 5 6
Mean Number of Contacts
High Communicators
All Other Staff
p <
Regular Informal Contact Outside of the Organization
0.05
p < 0.05
p < 0.05
p < 0.01
p < 0.01
p < 0.01
p < 0.001
The Gatekeeper as a Link to Outside
Technology
X X
X X
X
Literature
Outside Experts
Gatekeeper
Gatekeeper Characteristics
• High Technical Performance
• Not 'just communicators’
• Highest technical performers in the organization.
• Cannot be created by management.
• Low in the Organizational Hierarchy
• Concentrated at first level of technical supervision or below.
• Seldom found at higher levels of management.
• Seldom found on the technical ladder.
• Visibility
• They are easy to identify.
• Everyone knows who they are.
• Approachability
• Must be at least receptive to people.
International Gatekeepers
�International Gatekeepers tend to be Engineers or Scientists, who have worked in other countries and returned home.
�Engineers and Scientists visiting from other countries had very high foreign contact, but insufficient domestic contact to be International Gatekeepers.
A Managerial Career
VP
AD
DM
DH
SH $$$
GS $$
LE $
The Dual Ladder
TechnicalManagerial
LE
GS
SH
DH
DM
AD
VP
$
$$
$$$
?
SDSS $$$
SSS $$
SS $
Engineer A
Engineer B
Engineer C
Distribution of Positions in One Firm's Dual Ladder
0
10
20
30
40
50
60
70
80
90
100
Pro
po
rtio
n (
Perc
en
t)
Managerial Technical
The Dual Ladder
TechnicalManagerial
LE
GS
SH
DH
DM
AD
VP
$
$$
$$$ SDSS $$$
SSS $$
SS $
Engineer A
Engineer B
Engineer C
Criteria for Technical Ladder Promotion
TechnicalManagerial
LE
GS
SH
DH
DM
AD
VP
$
$$
$$$
SS
SSS
SDSS
$
$$
$$$
Engineer A
Engineer B
Engineer C
The Biggest Problem with the Dual Ladder
TechnicalManagerial
LE
GS
SH
DH
DM
AD
VP
$
$$
$$$
SS
SSS
SDSS
$
$$
$$$
Engineer A
Engineer B
Engineer C
The Dual Ladder
TechnicalManagerial
LE
GS
SH
DH
DM
AD
VP
$
$$
$$$
SS
SSS
SDSS
$
$$
$$$
Engineer A
Engineer B
Engineer C
Choosing Each of Three Possible Career Paths
Proportion of Engineers & Scientists in Ten Organizations
� MANAGEMENT 32%
� TECHNICAL LADDER 20%
� PROJECT ASSIGNMENT 48%
Career Preference as a Function of Age (N = 1,402)
0
10
20
30
40
50
60
70
20 30 40 50 60
Technical Ladder
Projects
Age (Years)
Pro
po
rtio
n
(Pe
rce
nt)
Management
Interesting
Career Preferences of Technical Ladder Staff as a Function of Age (N = 351)
Pro
po
rtio
n
(Pe
rce
nt)
70
Interesting
60 Projects
Management
50
40
30
20
Technical10
Ladder
0
20 30 40 50 60
Age (Years)
Career Preferences of Managers as a Function of
Age (N = 374)
0
20
40
60
80
100
120
20 30 40 50
Technical Ladder
Projects
60
Age (Years)
Pro
po
rtio
n
(Pe
rce
nt)
Management
Interesting
The Process of Innovation
INNOVATION
Technology
Market
Departmental Organization
D1 D6D2 D3 D4 D5
Technology
Market
Departmental Organization
D1 D2 D3 D4 D5
Technology
Market
Time & Coordination
Project Team Organization
Technology
P1 P6P2 P3 P4 P5
Market
Matrix Organization
Technology
Market
D1 D6D2 D3 D4 D5
P4 P5 P6 P7 P8P1 P2 P3
Matrix Organization
Technology
Market
D1 D6D2 D3 D4 D5
P1 P2 P3 P4 P5 P6 P7 P8
The Basic Tradeoff and Dilemma in Product Development Organization
• Departmental Organization • Project Team Organization• Departmental structure is more
closely mapped to the structure of the supporting technologies
• Project Team structure groups people from different disciplines together in a single team all reporting to a common manager.
• It thereby provides a better connection to those • It thereby provides better
technologies and better coordination of the project tasks and increased sensitivity to
ongoing technical support to market dynamics.
the project effort. • This is, however, accomplished
• This is, however, accomplished at the cost of much greater difficulty in coordination of the project tasks and less responsiveness to market change.
at the cost of a separation from the disciplinary knowledge underlying the project effort. When this is carried to an extreme, it will gradually erode the technology base of the organization.
Organizational Structure Space I
Iss
subsystem interdependenceIss =
rate of change of knowledge
dK dt =
dKdt
Organizational Structure Space II
x x x
x x x
o o o
o oo o
Project Team
Departments
dK dt
Iss
+
+
+
+
+
+
+
+ +
+
Organizational Structure Space III
Department
Project
Team
dK dt
Iss
Ti
T2
T1
T1 > T2
Structuring the Organization
• Standard Industrial Practice– Ignores the rate at which technologies are
developing (despite the fact that this can often be measured).
– Usually ignores the interdependencies in project work (seasoned project managers are an exception).
– Focuses on project duration (and usually makes the wrong decision on this parameter).
Organizational Structure Space IV
dM dt
Iss
Department
Project
Team
Ti
dKdt
Organizational Structure Space V
Department
Project
Team
Project
Team
Department dM dt
dK dt
Iss
Ti
BMW FIZ I Concurrent Engineering
Matrix Connections to Market and Technology
Technology
Dept
Head
Proj
Mgr
Dept
Head
Dept
Head
Dept
Head
Dept
Head
Dept
Head
Proj
Mgr
Proj
Mgr
Proj
Mgr
Proj
Mgr
Proj
Mgr Mark
et
Ma
rke
t
Matrix Connections to Market and Technology
Technology
Pr
Mgr
Dept
Head
Dept
Head Dept
Head
Dept
Head Dept
Head
IPT
Dept
Head
Dept
Head
IPT
IPT
Mgr IPT
Mgr IPT
Mgr IPT
Some Problems
Dept
Head
Dept
Head Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Technology
IPT
IPT
IPT
IPT
IPT
Mark
et
Competition for Resources
Problems with Imbalance
Dept
Head
Dept
Head Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Technology
IPT
IPT
IPT
IPT
IPT
Mark
et
OUT
Te
ch
In
teg
rity
The Need for Balance
Dept
Head
Dept
Head Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Technology
IPT
IPT
IPT
IPT
IPT
Mark
et
OUT
BALANCE
The Inescapable Conflict
Dept
Head
Dept
Head Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Head Dept
Technology
IPT
IPT
IPT
IPT
IPT
Mark
et
OUT
CONFLICT
A More Complete Matrix Using Integrated Product Teams
Dept
Head
Dept
Head Head Head Head Dept Dept Dept
PROD
DEV
Dept
Head
Dept
Head
Dept
Head
Dept
Head
Dept
Head IPT
IPT
IPT
IPT
IPT
MFGMKTG Integrated
Product
Teams
Matrix Connections to Product Development and
Manufacturing Engineering
Dept
HeadDept
Head
Dept
Head
Dept
HeadDept
Head
Product
Development
Dept
Head
Dept
Head Dept
Head
Manufacturing
Engineering
IPT
IPT
IPT
IPT
IPT
IPT
Integrated
Product
Teams
Management of Transitions
• The critical points of vulnerability in the life of a project are the points of transition. – Transitions can involve many parameters, for example:
• People
• Management
• Leadership & leadership style.
• Primary organizational responsibility and reporting relationships.
• Nature of the work.
• Types of knowledge required.• Physical location.
• To change all of these simultaneously is to court disaster.
Management of Transitions II
Transition Points P
roje
ct
Siz
e &
Sco
pe
Time
Management of Transitions IV
• Projects must be protected through transitions. – There must be areas of continuity to offset the areas of
change.
– Team size must grow in a gradual fashion.• This has implications for both organizational structure and
physical architecture.
• Both must be very flexible to allow this to happen along with a gradual transition in reporting relationship.
– There should be an extra effort to retain a sense of ‘ownership’ among team members.
• Avoid ‘runway management’.
Management of Transitions III
Pro
ject
Siz
e &
Sco
pe
Transition Points
Time
social order
Tom Allen, MITspatial order organizational order
Transition Performance
0
1
0.2
0.4
0.6
0.8
Separate Co-Located
Location of Engineers
Pe
rfo
rma
nc
e
p < 0.001
PERFORMANCE AS A FUNCTION OF
GROUP AGE
(PELZ & ANDREWS
40
45
50
55
0 2 4 6 8 10 12
Mean Tenure of Group Members (Years)
P e
rfo
rm a
nc
e P
erc
en
tile
Project Performance as a Function of Team
Age
(45 Chemical Industry Projects)
0
1
0 2 4 6
0.5
Mean Tenure of Project Team Members
(Years)
Sm
oo
the
d P
roje
ct
Pe
rfo
rma
nc
e
External Technical Communication as a Function of Team Age
(45 Chemical Industry Projects)
3.5
4
4.5
5
5.5
0 2 4 6 8
jec
t
Mean Tenure of Project Team Mem bers
(Years)
S m
o o
th e
d P
ro
P e
rfo
rm a
nc
e
TENURE
(PELZ & ANDREWS)
30
35
40
45
50
55
60
65
0 2 4 6 8 10 12
WORK PREFERENCES AS A FUNCTION OF MEAN
Pro
po
rtio
n o
f G
rou
p M
em
bers
Rep
ort
ing
(Perc
en
t)
PREFERENCE FOR BROAD
PREFERENCE FOR DEEP
Mean Tenure of Group Members (Years)
Perceived Influence Over Project Goals &
Objectives (Teams with Mean Tenure Greater Than Five Years)
Low
HighPro
ject
Team
Perf
orm
an
ce
Engineers Project Manager
0 0.2 0.4 0.6 0.8 1
Mean Perceived Influence
PERFORMANCE AS A FUNCTION OF GROUP AGE
(PELZ & ANDREWS)
40
45
50
55
0.5 2.5 4.5 6.5 8.5 10.5
Mean T enure of Group Members (Years)
Gro
up
Pe
rfo
rma
nc
e
Project Performance as a Function of Team Age
(45 Chemical Industry Projects) S
mo
oth
ed
Pe
rfo
rma
nc
e
0 2 4 6 8 10 12 14
Mean Tenure of Team Members (Years)
Project Performance as a Function of Team Age(45 Chemical Industry Projects)
Sm
oo
thed
Pe
rfo
rma
nc
e
0 2 4 6 8
Mean Tenure of Team Members (Years)
Project Performance and External Communication as a
Function of Team Age
(45 Chemical Industry Projects)
0
2
0 1 2 3 4 5 6
0.4
0.8
1.2
1.6
Mean Tenure of Team Members (Years)
Co
mm
un
ica
tio
ns
Pe
r
Pe
rso
n P
er
We
ek
Project Performance and External Communication as a
Function of Team Age
(45 Chemical Industry Projects)
0
1
0 2 4 6 8
0
2
0.2
0.4
0.6
0.8
Mean Tenure of Team Members (Years)
Sm
oo
the
d
Pe
rfo
rma
nc
e
0.4
0.8
1.2
1.6
Co
mm
un
ica
tio
ns
Pe
r
Pe
rso
n P
er
We
ek
TENURE
(PELZ & ANDREWS)
30
35
40
45
50
55
60
65
0 2 4 6 8
WORK PREFERENCES AS A FUNCTION OF MEAN
10 12 Mean Tenure of Group Members
(Years)
Deg
ree o
f P
refe
ren
ce
Broad Exploration of New Areas
Deep Exploration of Narrow Areas
Perceived Influence Over Project Goals & Objectives
(Teams with Mean Tenure Greater Than Five Years)
0 1
Low
0.2 0.4 0.6 0.8
High
Te
am
Pe
rfo
rma
nc
e
Degree of Influence
Project Manager
T eam Members
We Shape Our Buildings
"On the night of May 10, 1941, with one of the last bombs of the last
serious raid, our House of Commons was destroyed by the violence of the
enemy, and we have now to consider whether we should build it up
again,and how, and when. We shape our buildings, and afterwards our
buildings shape us. Having dwelt and served for more than forty years in
the late Chamber, and having derived very great pleasure and advantage
therefrom, I, naturally, should like to see it restored in all essentials to its
old form, convenience and dignity."
-WSC, 28 October 1943 to the House of Commons (meeting in the House
of Lords).
Notes: The old House of Commons was rebuilt in 1950 in its old form,
remaining insufficient to seat all its members. Churchill was against "giving
each member a desk to sit at and a lid to bang" because, he explained, the
House would be mostly empty most of the time; whereas, at critical votes and
moments, it would fill beyond capacity, with members spilling out into the
aisles, in his view a suitable "sense of crowd and urgency."
Distance Measurement
Distance
X
X
X
X
X
X
X
X
X
X
X X
X
X
X
Proportion of Communication Partners as a Function of
Probability of Technical Communication as a Function of P
ro b
a b
ilit
y o
f W
e e
k ly
Distance Between Work Stations
C o
m m
u n
ica
tio
n
0.25
0.2
0.15
0.1
0.05
0
0 20 40 60 80 100
Separation Distance (Meters)
Intradepartmental and Interdepartmental Communication and Physical Separation
0.4
0.3
0.2
0.1
0
0 20 40 60
INTRA-DEPARTMENTAL
INTER-DEPARTMENTAL
Sm
oo
the
d P
rob
ab
ilit
y o
f C
om
m u
nic
ati
on
Separation Distance (Meters)
The Effect of Organization I
DISTANCE
P(C)
D = f(1/N)
The Effect of Organization II
DISTANCE
P(C)
D = f(1/N)
P = f(Iss)
Some Obvious Points
P(C)
p1
p3
p2
S 1
S 2
S 3
<S2
S1
p2 p
1<
p3
p1>
<S3 S1
Distance
Departmental Size
1
0.8
0.6
0.4
0.2
0
0 10 20 30 40
Size of Department
Smoothed P(C)
Raw Data
Power (Raw Data)
Pro
ba
bil
ity
of
Co
mm
un
ica
tio
n
50
Face-to-Face and Telephone Communication
0.0001 0.0002 0.0005 0.001 0.002 0.005 0.01 0.02 0.050.0001
0.0002
0.0005
0.001
0.002
0.005
0.01
0.02
0.05
PROBABILITY OF FACE-TO-FACECOMMUNICATION
PR
OB
AB
ILIT
Y O
F T
EL
EP
HO
NE
CO
MM
UN
ICA
TIO
N
‘Bandwidth’ Limitation
Within a
Floor
Within a
Building
Within a
Site
Floor
Within a
Building
Site
Between
Sites
Low Complexity Information
Within a
Within a
Between
Sites
High Complexity Information
0 20 40 60 80 100 0 20 40 60 80 100
Proportion of ContactsProportion of Contacts
Face-to-Face Telephone Face-to-Face Telephone
Steelcase Ground Floor
Steelcase Third Floor
Effect of New Steelcase Building on Breadth of Communication
14
12
10
8
6
4
2
0
M ove to New Building
M e
an
Nu
m b
er
of
Co
m m
un
ica
tio
n P
art
ne
rs p
er
Pe
rso
n
0 2 4 6 8 10 12 14 16 18 20
Time (Weeks)
![Data-Driven Robust Optimization arXiv:1401.0212v2 [math.OC] 23 … · 2014-11-25 · Data-Driven Robust Optimization Dimitris Bertsimas Sloan School of Management, Massachusetts Institute](https://img.pdfslide.net/doc/110x75/5ed952e0f59b0f56f45f4683/data-driven-robust-optimization-arxiv14010212v2-mathoc-23-2014-11-25-data-driven.jpg)
