Sie sind auf Seite 1von 9

TOOL KIT

Are Your Engineers Talking to


One Another When They Should?
Cost overruns, schedule slippage, and quality problems often result from a failure to provide
timely information or resources. Here’s a way to help prevent that from happening.
by Manuel E. Sosa, Steven D. Eppinger, and Craig M. Rowles

C
OMPANIES THAT DESIGN COMPLEX, highly engineered prod-
ucts all have their horror stories. Ford and Bridgestone
Firestone lost billions of dollars after their failure to co-
ordinate the vehicle design of the Ford Explorer with the
design of its tires. Similarly, Airbus’s development of the A380
“superjumbo” suffered major delays and cost overruns because
of late emerging incompatibilities in the design of the electri-
cal harnesses of various sections of the plane’s fuselage. These
mistakes probably contributed to the loss of Airbus’s CEO and to
important changes in the management of the A380 program.
What’s striking about these stories and many others like them
Jude Buffum

is that in virtually every case, the people involved all agreed,


in hindsight, that they could have avoided their expensive mis-
takes by making sure that the different teams responsible for

hbr.org | November 2007 | Harvard Business Review 133

1657 Sosa.indd 133 10/5/07 7:30:04 PM


TOOL KIT | Are Your Engineers Talking to One Another When They Should?

developing the products’ components management tool: the design structure gine was divided into eight subsystems,
had communicated more effectively. Of matrix (DSM). Our application helps each of which was further decomposed
course with complex development proj- managers identify where failures in into five to ten components, for a total
ects, you can never be certain that you planned communications could occur of 54 engine components. Typical for
have planned for every contingency. as well as recognize when project teams such projects, the development organi-
However, our experience shows that engage in unplanned technical commu- zation was correspondingly structured
in the design phase of such projects, nications. We also analyze communica- into 54 cross-functional design teams
many companies would benefit from tions between project teams that take responsible for each component, plus
focusing sharply on the critical points place both within and outside the for- six integration teams responsible for
of contact among their various compo- mal project structure. We conclude by managing engine-level requirements
nent development teams to ensure that discussing how managers should han- in areas such as fuel efficiency. These
everyone knows when and with whom dle communications problems that are teams had to interact a lot: There were
they should be sharing information. revealed in the process. While we do not several hundred interfaces among the
This article helps managers mitigate pretend to offer a definitive solution engine components, many of which
this design communication problem. to the design communication problem, would have experienced significant
Drawing on detailed research into how we do believe that managers who use problems without proper communica-
Pratt & Whitney handled the develop- our tools over succeeding generations tion among the relevant design teams.
ment of its highest-thrust commercial of products will improve the quality of To help manage the communications
jet engine – the PW4098, which pow- their development processes. aspect of such projects, we propose the
ers the Boeing 777 – we present a new following approach: (a) identify unat-
application of an established project Catching Missed Interfaces tended interfaces, areas where commu-
Before They Occur nication should be occurring but is not,
Article at a Glance The first thing a project team does when and (b) look for unidentified interfaces,
faced with a complex development areas where communication is occur-
Companies that design complex prod- challenge is break the project down ring but has not been planned. The re-
ucts all have their horror stories. Yet, into manageable pieces that are then sult of implementing this approach is
they can all avoid mistakes by ensuring assigned to dedicated subteams. In the what we call an alignment matrix, which
that the different teams responsible for
context of developing a product like a reveals mismatches between the com-
developing the components of the prod-
jet engine, this results in a large number munications and exchanges that are
ucts communicate more effectively.
of specialized cross-functional teams, supposed to occur and those that actu-
A new application of an established each working on a different component ally do. It also demonstrates, therefore,
project management tool, the design or subsystem of the engine. Of course, how well the project has been planned
structure matrix, can help a company these teams cannot work in isolation; and executed.
identify where failures in planned in addition to designing their assigned To see how the approach works,
communications could occur as well as components, they must also integrate let’s suppose that we plan to develop
recognize when project teams engage their designs with those of the other a product with four components, each
in useful technical communications that
components to ensure that the entire of which requires its own specialized
were not planned.
product or system functions as a whole. design team. (This approach may be
If a company finds that a lot of planned It is critical, therefore, in planning a used when the organizational structure
communication is not taking place, then complex product development process maps directly to the product architec-
it should revisit its product develop- that the project managers specify just ture – that is, component X is designed
ment organization. Even projects that which resources and information differ- by team X – which is the case in most
are completely organized around a ent teams will need from each other at of the complex development projects
product’s architecture are typically vul- particular stages of the project. in the automobile and aerospace firms
nerable to communication breakdowns. In the Pratt & Whitney jet engine we have studied. For cases in which
development project we studied, the en- the organizational and product struc-
A company can ensure that critical com-
munication occurs by tasking special
teams (or the teams involved) with Manuel E. Sosa is an assistant professor of technology and operations management at
making sure that the right people talk to Insead in Fontainebleau, France. Steven D. Eppinger is a professor of management science
each other. It’s also important to ensure and engineering systems and deputy dean at MIT’s Sloan School of Management in Cam-
that the teams are working with the bridge, Massachusetts. Craig M. Rowles was a Module Center manager at Pratt & Whitney,
optimum communication tools. based in East Hartford, Connecticut; he is now CEO of Emegear, a medical device company
in Carpinteria, California.

134 Harvard Business Review | November 2007 | hbr.org

1657 Sosa.indd 134 10/5/07 7:30:13 PM


Assessing the Fit Between Design and Communication
The first step in analyzing the communication problem among In the design interface matrix below, a shaded cell indicates
design engineers is to have the system architects identify the presence of a technical interface between two components.
technical interfaces between components and record their Thus, the first row of the matrix indicates that component A will
responses on a design interface matrix. Next, have component require some input from components B and D but nothing from C.
and subsystem design team members identify the technical The first column shows that A will be expected to provide input to
interactions they have had, are having, or expect to have with (or impose technical constraints on) B and D but not C.
other teams, and present those responses on a team interaction In the team interaction matrix, each row indicates from which
matrix. Then merge the two matrices to generate an alignment other teams a particular team expects to need information and re-
matrix that reveals matches and mismatches between interfaces sources, and each column shows where a team will be expected
and interactions. The matrices below are based on a product with to provide the information and resources. A shaded cell indicates
four components. where the teams expect to interact.

Alignment Matrix Key


Design Interface Matrix
Providing components Matched interface: design
Component A depends interface that is matched by
A B C D
Receiving components

System on component D an actual or expected team


Architects’ A • interaction
Input
B • Unattended interface: design
identify technical Alignment Matrix
C interface identified by system
interfaces between • Providing component teams
architects that lacks corre-
Receiving component teams

components
D • A B C D sponding team interaction
A • Unidentified interface: team
B • interaction that takes place
Team Interaction Matrix or is expected to take place
Providing teams C • without a corresponding design
interface identified by system
Design A B C D D •
architects
Receiving teams

Teams’ A •
Input Lack of interdependence:
determine technical B • components that do not share
interactions teams an interface or involve design
C •
have had, are having, team interactions
or expect to have D •
Team C requests
information from Not applicable
team D •

tures do not map directly, refer to M.E. transfer forces, material, energy, or in- people working on each component
Sosa’s “Aligning Process, Product, and formation to other components to en- design team expect to take place with
Organizational Architectures in Soft- able them to work properly? Answers people from the other teams. Specifi-
ware Development” in the Proceedings to such questions are used to identify cally, we ask members of each team
of the 14th International Conference in the interfaces among all the compo- whether they anticipate the need for
Engineering Design, Paris, August 2007.) nents of the product. technical information or resources
Our analysis involves the following Armed with this data, the project from other teams. In surveying them,
three steps: managers can present the responses we need to make sure they are all fa-
1. Interview the system architects. on a four-by-four design interface ma- miliar with the function and specifica-
We begin by requiring the senior en- trix (a type of DSM used to map the tions of the component or components
gineers responsible for the product’s network of component interfaces), they are developing. (We do not share
overall function and layout – the sys- such as that illustrated in the exhibit with them the matrix produced in the
tem architects in engineering lingo – to “Assessing the Fit Between Design and first step, as this would bias the teams’
identify the technical design interfaces Communication.” responses.) Using these survey data, we
among the four components. Do com- 2. Survey the component design can represent the technical interaction
ponents need to be spatially connected teams. In the second step, we identify patterns of the teams in another four-
with each other? Do some components the technical communications that the by-four matrix (corresponding to the

hbr.org | November 2007 | Harvard Business Review 135

1657 Sosa.indd 135 10/5/07 7:30:18 PM


TOOL KIT | Are Your Engineers Talking to One Another When They Should?

four design teams) that we call a team state where they actually received the (vibration or heat), structural forces, or
interaction matrix, also shown in the ex- information and resources they needed. signals used by the control system of
hibit “Assessing the Fit Between Design Overlaying the new team interaction the engine. The design interface matrix
and Communication.” matrix on the original design inter- drawn from this information is shown
3. Combine the results. In the third face matrix would reveal whether the in the exhibit “Creating an Alignment
and final step, we overlay the two ma- mismatches uncovered at the start of Matrix for Pratt & Whitney.” We then
trices to obtain the alignment matrix the project had persisted and whether asked at least two members of each
(again, see the exhibit “Assessing the other mismatches had appeared. A of the 54 teams involved in the proj-
Fit Between Design and Communica- postmortem of this kind yields valuable ect how often they received technical
tion). This matrix shows the matches insights into future product develop- information from other teams during
and mismatches between the product’s ment projects, especially for companies the detailed design phase of the proj-
ect and how critical they perceived this
information to be. The results of this
survey documented 423 interactions
among component teams, which ap-
pear on the team interaction matrix
shown in the exhibit. We finally com-
puted an alignment matrix by merging
the two first matrices.
Readers armed with a magnifying
glass would count 220 unattended de-
sign interfaces that were not matched
by team interactions and 74 uniden-
tified interfaces in which teams ex-
changed technical information even
though there was no identified design
interface between the components. Al-
though it would be naive to expect a
perfect alignment of design interfaces
and team interactions – and, in this
case, many of the 220 unattended in-
terfaces were unproblematic or not
critical – misalignment on this scale
indicates that Pratt & Whitney was sub-
architecture as conceived by the system that expect to develop similar products ject to considerable risks involving cost
architects and the expectations of the in the future or further generations of overruns or other problems with this
teams involved in the product’s devel- the same product. project.
opment. More important, it highlights We conducted one such analysis of
the mismatches – when design inter- Pratt & Whitney’s development of the Why Mismatches Occur
faces have been identified but team PW4098, the engine that, at the time, Mismatches do not occur randomly
interactions are not taking place (un- set new standards in the aviation indus- in a product or organization. Rather,
attended interfaces) and when team try for development speed and cost. (It they are the result of product design
interactions take place without a corre- was also the first commercial jet engine and organizational factors. Planned
sponding design interface being identi- ever certified for 180-minute extended- key communication points may remain
fied by system architects (unidentified range twin operations from its first day problematic for several reasons, in-
interfaces). As we shall see, sometimes of service.) We began by interviewing cluding the presence of organizational
these unidentified interfaces turn out the engine’s architects, who identified boundaries (cross-boundary interfaces
to be critical. 569 interfaces among the 54 main com- are more likely to be missed than in-
At the end of a project, managers ponents of the engine. Many of these terfaces with a team belonging to the
can update the alignment matrix by interfaces were critical and complex be- same group), the lack of interface criti-
redrawing the team interaction matrix cause they not only involved physically cality (complex and critical interfaces
based on more recent surveys in which adjacent components but also the trans- receive more attention than noncritical
component team members are asked to fer of material (air, fuel, or oil), energy ones), the use of indirect communica-

136 Harvard Business Review | November 2007 | hbr.org

1657 Sosa.indd 136 10/5/07 7:30:25 PM


TOOL KIT | Are Your Engineers Talking to One Another When They Should?

Creating an Alignment Matrix for Pratt & Whitney


This exhibit shows the design interface, team interaction, and project. Note that components (in the design interface matrix)
alignment matrices we developed for Pratt & Whitney after the and teams (in the team interaction matrix) that belong to the
PW4098 project. It is striking how many mismatches (294 in all) same subsystem are clustered together so that we can easily
turned up even in a post hoc analysis. Conducting this exercise distinguish (with the boxes along the diagonal) between inter-
before or during the project would probably have revealed many faces (and interactions) that occur within boundaries versus
additional mismatches that got fixed during the course of the those that fall across boundaries.

Design Interface Matrix


System engineers identified
1 0 0 0 0 0 0 0 0 0 6 3 0 3 0 0 0 0 0 0 0
6 0 0 0 0 0 0 0 0 5 5 6 6 6 3 3 0 0 2 0 0 0
6 0 6 4 4 8 4 4 0 1 0 0 4
6 0 4 4 4 0 0 2 0 0 0 0
4 4 0 0 0 0 0 0 0 0 0 0

2
3
4
8
6
6
4
4
2
1
2
0
0
0
0
4
0
0
4
0
4
4
0
4
0
6
0
4
6
0
0
0
6
6
0
0
4
4
0
0
0
0
0
0
8
8
0
0
6
6
5
5
2
2
5
5
2
2
2
2
0
0
0
0
0
0
569 design interfaces among
engine components
0 6 3 1 6 6 0 0 0 0 0 0 0 0 0 0 0 0 0
2 5 2 0 0 4 4 0 0 6 0 0 0 0 0 0 0 0 0 0 0
4 5 0 0 0 0 0 0 6 0 4 6 0 0 0 0 0 0 0 0
0 0 0 0 0 8 8 0 0 3 0 4 0 0 0 0 0 0 0 0

Alignment Matrix Key


6 0 0 0 6 6 6 0 0 6 4 0 6 6 6 2 2 0 0 8
0 2 0 0 0 2 2 1 6 0 2 5 5 4 8 7
0 0 0 0 0 0 0 0 4 2 0 7 7 0 3 0
0 2 0 0 0 2 2 1 5 5 7 0 8 2 2 0
1 1 1 0 0 2 2 1 2 5 7 8 0 6 2 1
1 1 0 0 0 0 0 2 4 0 2 6 0 0 0
4 8 3 2 2 0 0 8
0 7 0 0 1 0 8 0
0 0 6 4 0 6 0 8 0 0 0
0
0
0
0
6
4
0
4
0
7
5
8
5
0
0
0
5
2
0
0
5
0
0
0
0
0
0
0
4
4
0
0
0
0
0
0
0
4
0
0
4
0
0
0
Matched interface:
0
0
0
0
0
5
0
0
0
2
0
0
4
5
0
2
0
0
0
0
0
4
0
0
0
4
4
7
4
0
4
2
4
4
0
6
7
2
5
0
4
6
6
2
0
0
0
2
0
0
2
0
0
0
0
0
4
0
0
0
0
0
0
0
2
0
0
2
0
0
0
4
0
0
0
(349 instances) design
0 0 0 0 6 0 0 0 0 0 0 0 0
0
0
6
0
7
0
0
0
0
0
0
0
2
0
0
0
6
0
0
0
6
0
1
0
4
4
0
0
0
0
6
0
0
0
0
2
7
0
0
6
0
0
4
0
4
0
7
5
0
8
3
5
4
0
7
3
0
6
0
0
0
0
0 0 0 0 0 4
0 0
0
0
0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Alignment Matrix interface that is matched by
0 2 0 0 0 7 0 0 6 3

0
0
6
4
4
0
6
0
0
0
0
3
4
0
0
5
0
0
7
0
4
0
2
7
0
0
3
3
6
0
7 0
4
5
0
0
3
3
0
3
6
3
6
6
0
0
6
2
2
2
4
0
0
2
2
2
2
0
4
0
0
0
3
2
2
0
4
0
0
0
0

0
0
0
1
0
0
0
0
0

1
0
0
1
0
0
0

0
0
1

0
an actual team interaction
0 6 0 3 0 0 0 0 0 0 4 0 0 2 0 2 2 4 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 6 0 0 0 0 2 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 2 0 0 6 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 6 0 0 0 0 2 2 0 0 0 0 0 3 2 0 4 0 0 0 0 1 0 1 1 1 1 0
0 6 0 2 0 0 0 0 0 2 2 2 2 2 2 2 0 4 0 2 4 2 0 0 2 0 0 0 0 0 0 0 1 1 1 1 1 0 0
3
0
3
0
3
6
3
3
0 0
0
0
0
0
2
0
0
6
6
4
0
4
0
5
5
0
0
0
0
0
0
0
0
6
0
0
0
7
0
5
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
8
0
5
0
0
0
0
0
6
0
0
0
6
0
0
0
0
0
0
0
6
0
6
0
0
0
0
0
6
0
8
0
8
0
8
5
0
0
0
0
6
0
0
7
5
0
7
0
7
0
7
7
0
0
0
0
0
0
6
0
6
0
4
4
0
6
4
2
0
0
0
0
6
2
0
6
6
0
2
0
6
6
2
6
0
0
6
0
0
6
6
6
0
0
6
4
8
6
6
8
8
6
8
6 0
0

0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0 0
0
0
0
0
0
0
1
0
0
1
0
0
0

0
0
0
Unattended interface:
(220 instances) design
0 0 0 4 5 0 0 4 3 0 0 0 0 0 0 0 7 0 0 0 0 0 0 6 0 0 0 0 0 4 5 6 6 0 0 6 4 7 7 0 0 0 0 0 1 1 0 0 1 0 1 1 1 1 0
0 0 0 0 0 0 0 0 0 0 0 0 6 4 0 0 4 0 0 0 0 0 0 3 0 0 0 0 0 0 0 6 0 0 0 6 0 2 6 0 0 0 0 0 0 0 0 0 1 1 0 1 1 0 0 0 0 0 0 0 0 0 0
4 2 0 6 6 0 0 6 2 0 0 0 8 8 0 0 0 0 7 0 7 4 0 7 0 0 0 7 9 4 5 6 6 6 5 0 8 8 8 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 6 0 0 0 2 2 6 0 0 0 0 0 0 0 0 4 0 0 0 0 0 4 0 6 0 0 9 0 4 0 0 0 5 7 7 0 6 4 4 0 6 0 0 6 0 0 0 0 0 0 1 0 0 0 1 0 1 0 1 1 1 0 0 0 0
0 0 0 0 0 0 0 0 0 4 3 0 0 8 6 0 0 0 0 4 0 0 4 0 0 0 0 0 0 0 0 0 6 6 0 0 0 0 8 6 6 6 6 0 0 0 0 4 1 1 0 0 0 0 1 0 0 0 1 0 1 0 1 1 0 0 0 0 0
4 2 0 0 0 0 0 0 0 5 3 0 0 4 0 0 0 0 2 0 0 0 2 0 0 0 0 8 0 8 8 0 3 0 0 0 6 0 8 4 6 6 6 4 6 6 4 0 0 0 0 0 0 0 0 0 0 0 1

2
0
0 0 1
0

0
0
1
0
1
0
0 0 0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
interface identified by system
architects that lacks corre-
0 0 0 0 0 0 0 0 0 1
0 0 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1
1 1 0 0 0 0 1 1 0 0 0 1 0 1 1 0 1 0 0 0 0 0 0 0

1 0
1
0
0
0
1
0
1 0
0 1 1
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
0
0
0
0
0 0
0
0
0
0
sponding team interaction
Team Interaction Matrix 0 0 1 0 0
0
0
0
1
0
0
0
0

0
0
0 0

0
0

1
1

0
0 0

0
0
0
0
0
0
0
0 0 0
0

0 1 1
0 0 1 0 0 1
0
0 1 0
0
0
0 1
0
0 1 1 1
0
0
0
1

1
0

0
0
0
0
0

0 0 0
0

0
Unidentified interface:
0 1 1
0
1
0

1
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
1
1

1
0
1
0

1
0
0
0
0
0
0
0
1
0
0
0
1
0
0
0
1
0
0
0
0
0
0
1
0
0
0
1
1
0
0
0
0
0

0
(74 instances) team interac-
1 1 1 1 1 1 0 1 1 1 0 0 1 0 0 0 0 0 1 1

1
1
1
1
1 1 1 1
0
0
0
0
0
0
0

0
0
0
1
1
0
0
0
0
0
0

0
0
0

0
1 0

0
0
0

0
0
0

0
0
0

1 1
tion that takes place without a
corresponding design interface
0 0 0 0 0 0 0 0 0 0 0 0 0 1
1 1 1 1 1 1 0 0 0 0 0
1 0 0 0 0 0 0 0 3 0 0
1 1 1 1 0 1 0 0 0 0 0 0
1 0 0 0 0

identified by system architects

0
0
Lack of interdependence:
(2,219 instances) components
that do not share an interface or
0

0
Design teams reported the involve design team interactions
occurrence of 423 technical
0

interactions among them

For details, refer to M.E. Sosa, S.D. Eppinger, and C.M. Rowles, “The Misalignment of Product Architecture and Organizational Structure in Complex Product Development,”
Management Science 50, no. 12 (2004): 1674–1689.

tion channels (teams sometimes pass compressor, for example, were virtu- involved teams from different parts
technical information through other ally unchanged from a previous en- of the organization that had fewer
intermediary teams rather than inter- gine generation. As a result, some of opportunities for informal communi-
act directly), the presence of interface the teams responsible for these com- cations, which might have uncovered
carryover (interfaces that have been de- ponents had to pay only marginal at- changes to the previous standards. The
fined in previous projects may not need tention to coordinate their interfaces impact of these unattended interfaces
to be reconfirmed in designing the new among themselves, though they still probably resulted in very small reduc-
project), and the use of interface stan- needed to coordinate with other highly tions in performance or durability of
dardization (interfaces whose specifica- redesigned components. Such inatten- affected components and systems. But
tions have been formally documented tion trickled over into their communi- given the 25- to 30-year life expectancy
are supposed to remain unchanged). cations involving other highly designed of a product like the PW4098, even
In the case of the Pratt &Whitney components. In addition, many struc- small performance deviations could
project, some mismatches occurred be- tural and thermal design interfaces add up and cause significant warranty
cause component designs were carried were left unattended because they or service expense over the life of the
over from previous designs. A number were noncritical or assumed to be stan- product. For example, if a critical com-
of components of the high-pressure dard. In many cases, these interfaces ponent like a turbine airfoil misses its

138 Harvard Business Review | November 2007 | hbr.org

1657 Sosa.indd 138 10/5/07 7:30:31 PM


CORPORATE LEARNING
SOLUTIONS FOCUSED ON
YOUR SUCCESS CORPORATE AND EXECUTIVE EDUCATION life expectancy by only a few percent,
CUSTOM SOLUTIONS integrate industry’s it could mean one or more unplanned
best practices and Drexel University’s engine removals for maintenance over
LeBow College’s innovative development the life of the engine. For such a prod-
programs with your business culture and uct, a single unplanned engine removal
organizational capabilities. could add up to a $150,000 incremen-
LeBow’s Global Knowledge Network, tal cost to the customer.
comprised of leading researchers and Although most unattended inter-
sought-after business consultants, help faces are not critical, some are. Those
you to find contemporary ways to create critical unattended interfaces mostly
long-term sustainable value for your
occur when the teams involved come
company in the global marketplace.
from different parts of the organiza-
Programs are offered online or face-to- tion. The costs can be substantial. In the
face on Drexel’s Philadelphia campus or
on-site at your location. PW4098 project, two unattended inter-
faces turned out to be critical, and their
For more information on custom designing costs varied with the time it took for
a program to complement your organization’s
problems associated with the interfaces
strategic corporate training initiatives, call
215-895-1604, email executive@drexel.edu to be uncovered and resolved. One re-
or visit www.lebow.drexel.edu/execed lated to a change in the structural loads
transferred between rotating hardware
assemblies in the high-pressure core
COLLEGE OF BUSINESS and resulted in excessive loads on cou-
pling hardware during tests of early
LEARN HERE, LEAD ANYWHERE® LeBow development engines. Consequently,
Pratt & Whitney had to disassemble, re-
design, and rebuild the test engines. We
estimate that this one problem added
1% to 2% to the cost of the program and

Harvard Business Review a three- to four-month delay to certain


elements of the program. The other
unattended interface was uncovered
FOR THE BLIND later, after early production assembly
had begun but before any engines were
Subscriptions to Harvard Business Review are available in a special cassette shipped. This one, also related to loads
transferred between engine modules,
format at a new low price of $49 per year for individuals who are print handicapped. reduced the life expectancy of one of
the main engine bearings. The number
For further information on HBR for the Blind and other custom recordings services, of affected parts and engines was sig-
please contact: MAB Recording Studio, 313 Pleasant Street, Watertown, MA, nificantly greater than those associated
with the first problem – and so was the
at 617-972-9117 or e-mail rpierson@mabcommunity.org. impact on the program in terms of cost
and time.
We have found unidentified inter-
faces to be less common than unat-
tended interfaces. However, unlike un-
attended interfaces, their occurrence is
almost always positive for the project
because they reveal potentially critical
but unanticipated interdependencies
among a product’s parts or systems.
Many of the unidentified interfaces we
uncovered at Pratt & Whitney related
to investigations into possible engine-

1657 Sosa.indd 140 9/28/07 7:29:49 PM


A Call to Action
level design problems that could have Potential solutions can be varied, in-
resulted in excessive strain, overheat- cluding redrawing organizational lines,
ing, or insufficient pressure in the test reassigning or creating new interface
engine. Because the teams involved management responsibilities and fa-
talked to each other as they began to see cilitation tools, or even redesigning the
or anticipate the unexpected problems, system architecture (some of these are
they were able to mitigate them before discussed below). To find the solution
the product testing phase of the proj- appropriate for your project, consider
ect, resulting in considerable time and the following three steps:
cost savings potential. When unidenti- 1. Review organizational and system
fied interfaces like these are uncovered, boundaries. For projects in which a sig-
the main question is whether to for- nificant number of unattended inter-
mally incorporate them into a project’s faces span organizational boundaries,
schedule and routines or
leave them be. This deci-
sion depends largely on Most integration teams pay only
“Finally, somebody has written a book about
how critical the commu- marginal attention to the quality of what it really means to be a leader.”
nication is. In the case of Lydia Thomas, President and CEO, Noblis, Inc.
communication between component
the interfaces described
teams. That needs to change.
above, Pratt & Whitney
formalized some of the
relevant team interactions in planning project managers should revisit their or-
the development of its next-generation ganizational structures. Doing so would
engine. probably have helped Airbus avoid the
The conditions that generate un- problems it encountered. The company
identified interfaces are different from based the organizational structure of
those that cause people to overlook in- its programs not only on the architec-
terfaces. Several of Pratt & Whitney’s ture of the plane but also on the share
unidentified interfaces occurred be- of work owned by the various partners
tween teams working on engine-level of EADS (the European consortium to
design scenarios that created adverse which Airbus belongs). The addition of
structural or thermal loads. This, in this extra set of boundaries increased “Great, rich insights into how to improve
turn, generated the need for technical the likelihood of unattended interfaces succession planning.”
Steve McMillan, CEO, Stryker
solutions across distinct components. occurring and reduced the likelihood
In one case, three teams from both of problem-solving unidentified inter-
the high-pressure turbine and the low- faces taking place.
pressure turbine had to interact with To change the organizational struc-
teams working on the combustion ture, though, may necessitate changing
chamber to optimize the thermal en- the system architecture, because that
vironment and resulting durability of is what drives the organizational struc-
their respective components. These ture at most companies. Developing
were deemed critical interfaces that more-modular components that share
had not been identified by the system fewer direct and indirect interfaces
architects. Fortunately, the people on with other components in the product
these teams had worked together in the is especially helpful because technical
same roles in previous projects, making communication in such projects is eas-
it more likely that they would have un- ier to manage than in projects requir-
planned exchanges of information. ing a great deal of component interac- “Tabrizi’s core insights offer important
lessons not only for the business world,
tion. In the Pratt & Whitney project, but also for organizations in general.”
How to Manage Mismatches teams responsible for designing more- Eric Schmidt, Chairman and CEO, Google Inc.
Once the root causes of mismatches modular components missed far fewer
are understood, an organization can critical interfaces with teams from Available wherever books are sold

then decide how to deal with them. other components. There were fewer to
www.HBSPress.org

1657 Sosa.indd 141 9/28/07 7:30:04 PM


TOOL KIT | Are Your Engineers Talking to One Another When They Should?

be missed, and the workload associated properly attended to. In the PW4098 to complete work, the various tools and
with fewer direct and indirect inter- project, the secondary airflow team platforms must be carefully integrated.
faces was more predictable. Too much (one of the six integration teams) was In planning the development of the en-
modularity, though, can lead to myopia, responsible for managing the engine’s gine project that followed the PW4098,
particularly at the subcomponent level. multiple internal thermal and pres- for example, Pratt & Whitney linked full
At Pratt & Whitney we found that the sure management systems to optimize engine aerodynamics and secondary air
design teams of highly modularized engine aerodynamic performance and flow analytical models with component
subsystems were less likely to talk to component durability. It regularly set models to help the design teams man-
teams working on other modularized up meetings and other communica- age their interfaces with the support of
subsystems than were teams working tions between otherwise unconnected the integration teams. Specific people
on more integrated subsystems. In go- teams to address critical interfaces. on each team were then tasked with
ing modular, therefore, product de- Unfortunately, most integration tracking the impact of design changes
signers still need to pay careful atten- teams focus on milestone planning across the component and system mod-
tion to critical interfaces across those and resource allocation, paying only els and communicating those findings
subsystems. marginal attention to the quality of to their counterparts on other teams.
2. Form teams to handle misman- communication between component •••
aged interfaces. Managers also have the teams. That needs to change. Consider Our approach provides a systematic
option to manage critical interfaces – the pain that could have been avoided method for an organization to learn
to ensure that unidentified ones occur in the last phases of the development how and where it is exposed to the risk
or that unattended ones are formal- of the Airbus A380 if one of the integra- of communication failures between
ized – by assigning such work to the tion teams had realized that the elec- design teams working together to de-
teams already tasked with the inter- trical harnesses team in Germany and velop complex products. Moreover, an
action or by making people on the in-
volved teams formally and actively ac-
countable for the interfaces. We would Many design teams miss interfaces because project
recommend this approach for manag- planners haven’t thought through their use of
ing identified interfaces across bound- communication tools and shared platforms.
aries – the interfaces design teams are
most likely to ignore.
Another way to handle missed inter- its counterpart team in France, which organization can use our tools to deter-
faces – one that also avoids the need to were responsible for different sections mine how changes in system architec-
significantly change the organizational of the fuselage, were not properly com- ture, or the emergence and removal of
structure – is to extend the responsibil- municating about their design interface interfaces between system components,
ity of existing integration teams. Most specifications. will affect its ability to avoid communi-
large projects have such teams: At Pratt 3. Select appropriate communica- cation failures in the future. By using
& Whitney there were six teams man- tion support tools. Many design teams DSMs to document the architecture of
aging system issues like air and fuel ef- miss interfaces because project planners the product for every generation of a
ficiency, which affected the design of haven’t thought through their use of product family, managers can identify
practically all engine components. Even communication tools and shared plat- key differences between old and new ar-
though the management of team com- forms. At Airbus, one of the main rea- chitectures. With the alignment matrix,
munications is not usually the primary sons for the communication breakdown managers create a compact and visual
function of integration teams, by the between the A380 teams was the lack representation of interfaces and interac-
nature of their work, these teams com- of compatibility between the computer- tions that allows them to diagnose how
municate with almost all other teams aided design (CAD) tools they used. their organization addresses design in-
in the organization. Accordingly, they Being smarter about using communi- terfaces. Most important, the alignment
are in a position to learn in real time cation tools doesn’t have to involve a lot matrix can help managers properly di-
about the status of critical interfaces of technology: Pratt & Whitney requires rect their efforts to align team interac-
during the design process and to bring teams to regularly complete controlled tions with design interfaces to prevent
unconnected teams together to handle interface documents and component costly problems from occurring later in
critical interfaces that need special at- requirements documents (specifica- the product life cycle.
tention. These integration teams could tions) to ensure that critical interfaces
be made responsible for flagging those are identified and attended to. In cases Reprint R0711J
critical interfaces that are not being where team members use technology To order, see page 155.

142 Harvard Business Review | November 2007 | hbr.org

1657 Sosa.indd 142 10/5/07 7:30:47 PM