Modeling and Analysis of Integration
Processes for Engineering Systems
Meir Tahan and Joseph Z. Ben-Asher (2004)
Student 8
Jon Gibbs
16.842
16.842
Page 1
Types of Integration
• Single Stage
– Small number of subsystems
– Assumes subsystems acquired simultaneously
• Incremental
– System too complex for single stage
– Subsystems can arrive independently
– Multiple methods
– Typically have associated time and cost benefits
16.842
Page 2
Large Scale System Development Issues
• Subsystem Development
– Synchronizing Tasks
• System Development
– Integration of Subsystems
• System Performance
– Verification
16.842
Page 3
Relationship Strengths
4 Types of Flow: Information, Energy, Material, Physical
• Degree 0: No significant relationship
• Degree 1: One type of flow only
• Degree 2: Significant flow of two types
• Degree 3: Significant flow of three types
Sharman and Yassine [2004]
16.842
Page 4
Case Study - UAV
•
Four UAV subsystems
1. Engine
2. Flight control
3. Optronic payload
4. Payload control
•
How should these systems be integrated?
16.842
Page 5
Single Stage Integration
Single-stage integration.
1
Act1
0
2
0
Act2
0
Act3
0
7
d*Int12
8
d*Int34
9
Int1234
10
3
0
Act4
4
Image by MIT OpenCourseWare. Adapted from Figure 2
on p. 65 in Tahan, Meir, and Joseph Z. Ben-Asher.
“Modeling and Analysis of Integration Processes for
Engineering Systems.” Systems Engineering 8, no. 1
(2005): 62-77.
• Subsystems prepared for integration
• Tightly coupled subsystems integrated separately
• Final systems integration of integrated subsystems
• Total Time: Max(t1, t2, t3, t4) + d*t1,2 + d*t3,4 + t1,2,3,4
16.842
Page 6
Incremental Integration
Incremental integration.
Act1
Act2
0
5
0
Act3
Act4
6
0
q,Int34
8
d*Int12
9
S
0
p,Int12
0
7
Int1234
10
d*Int34
q=1-p
Image by MIT OpenCourseWare. Adapted from Figure 3 on p. 65 in Tahan, Meir, and Joseph Z. Ben-Asher.
“Modeling and Analysis of Integration Processes for Engineering Systems.” Systems Engineering 8, no. 1
(2005): 62-77.
• Subsystems prepared for integration
• First set of tightly coupled subsystems ready for integration triggers integration start
• Assumes that remaining two subsystems ready after first set is integrated
• Integration of all subsystems then performed
• Total Time: Min of [Max(t1,t2), Max(t3,t4)] + t3,4 + d*t1,2 + t1,2,3,4
16.842
Page 7
Success Probability
Summary of the probabilities of sucess in meeting the time requirement
Case 1
Case study
Case 2
d
0.8
1.0
1.2
0.8
1.0
1.2
Single stage
1.000
1.000
0.257
1.000
0.770
0.116
Incremental
1.000
1.000
0.949
1.000
0.998
0.853
Image by MIT OpenCourseWare. Adapted from Table 2 on p. 68 in Tahan, Meir, and Joseph Z. Ben-Asher.
“Modeling and Analysis of Integration Processes for Engineering Systems.” Systems Engineering 8, no. 1
(2005): 62-77.
• Incremental integration has higher probability of success
• Single stage integration will perform poorly with complex systems
16.842
Page 8
Types of Incremental Integration
• Option A: Single Stage
Minimize Total Integration Time
– Ts=Max(t1,t2,t3,t4)
• Option B: Integrate first available units:
t1
– Ts =Min[Max(t1,t2),Max(t3,t4)]
Then integrate other units in order of
t2
5
0
t3
6
0
0
q,t34
8
d*t12
9
S
p,t12
7
t1234
10
d*t34
t4
arrival
Option C: Integration in pairs followed by full
Image by MIT OpenCourseWare. Adapted from Figure 4 on p. 66 in Tahan, Meir, and
Joseph Z. Ben-Asher. “Modeling and Analysis of Integration Processes for Engineering
Systems.” Systems Engineering 8, no. 1 (2005): 62-77.
system integration
Option D: Integrate the first pair then integrate
all available units
16.842
Page 9
Incremental Options C & D
• Option C
• Option D
• (1) Integrate first available
• (1) Assume a simulator for unit 1
– Min[Max (t1,t2),Max (t3,t4)]
• (2) Integrate in pairs
– (t1,2,t3,4),
• (3) Full systems integration
– (t1,2,3,4)
• Ttot=Min[Max (t1,t2),Max
(t3,t4)]+t3,4+t2,3+t1,2,3,4
– Min[t2,Max (t3, t4)]
• (2) Integrate remaining pairs with
some difficulty
– d*t3,4 or d*t1,2
• (3) Full systems integration
• Ttot=Min[t2,(Max
(t3,t4)+t1,2]+d*t3,4+t2,3+t1,2,3,4
16.842
Page 10
Results
• Difficulty Factor mainly
determines break even point
0.1
Tot 0.05
.43
• Replacement time of simulator can
be traded with difficulty factor to
decide which method has the
fastest expected time
.23
0
145
150
155
160
w
B
165
170
175
D
Image by MIT OpenCourseWare. Adapted from Figure 10 on p. 73 in Tahan, Meir,
and Joseph Z. Ben-Asher. “Modeling and Analysis of Integration Processes for
Engineering Systems.” Systems Engineering 8, no. 1 (2005): 62-77.
P success [%]
50
40
Option D
30
20
Option B
10
0
0
5
10
15
20
25
X[%]
• Many firms know the replacement
time of a simulator with more
certainty than component
completion times
Image by MIT OpenCourseWare. Adapted from Figure 11 on p. 74 in Tahan, Meir,
and Joseph Z. Ben-Asher. “Modeling and Analysis of Integration Processes
for Engineering Systems.” Systems Engineering 8, no. 1 (2005): 62-77.
16.842
Page 11
Decision Considerations
• Single Vs. Incremental
– Consider level of complexity
• Simulator:
– Consider replacement time and
level of uncertainty
– Cost of test equipment and
labor
• Type of incremental integration
– Knowledge of PDF for each
method
16.842
Page 12
Very Light Jet Case Study
• Is it faster to prototype or not?
• Approach 1) Obtain FAA approval
first, make changes, tackle
manufacturing learning curve after
• Approach 2) Prototype first and
tackle manufacturing learning
curve, obtain FAA approval
second, make certification changes
last
• Three Variables: Tlc,Tfaa,d*Trework
16.842
Page 13
Conclusions
• Incremental integration almost always requires less time than single stage
integration
• Incremental integration may be required for complex systems
• Multiple levels of incremental integration exist
16.842
Page 14
Integration Example – Boeing 787/GEnX
• Six Primary Modules
• Fan, LPC, HPC, Combustor, HPT, LPT
Image by MIT OpenCourseWare.
16.842
Page 15
Boeing 787/GEnX Integration
• Aircraft and engines are typically developed in parallel at separate
companies and with minimal final integration (relative to
aircraft/engine design). What allows this incremental integration to
occur?
• Aircraft engine design is typically divided into modules
(subsystems). One could argue that the relationship between
modules is both tightly and weakly coupled. Why?
16.842
Page 16
Discussion Questions
• Many of the incremental integration models do not assume a
particular path, but rather perform sub-system and system integration
when the inputs are available. Is there a benefit from planning a
target path for integration?
• The paper presents a decent back-to-back study of single stage and
incremental integration through various modeling techniques. Is
there a benefit for a program to evaluate the integration methodology
using these techniques? If so, would the result be quantitative or just
qualitative?
• What level of simulator accuracy is required for incremental
integration?
16.842
Page 17
MIT OpenCourseWare
http://ocw.mit.edu
16.842 Fundamentals of Systems Engineering
Fall 2009
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.