16.842
Fundamentals of Systems Engineering
Integrated Lifecycle Management
and Modeling
4 December 2009
Prof. Olivier de Weck
1
Massachusetts Institute of Technology - Prof. de Weck
V-Model – Dec 4, 2009
Stakeholder
Analysis
Systems Engineering
Overview
Requirements
Definition
Commissioning
Operations
Cost and Schedule
Management
System Architecture
Concept Generation
Human
Factors
Lifecycle
Management
Tradespace Exploration
Concept Selection
Verification and
Validation
System Integration
Interface Management
Design Definition
Multidisciplinary Optimization
System
Safety
Today’s Topics
• Lifecycle Management
– First part: Conceive and Design
– Second part: Implement and Operate
• Lifecycle Modeling and Process
– What to model across lifecycle?
– Value Modeling and Optimization framework
• Summary and last Announcements
3
Lifecycle Management
modeling simulation
experiments
The Enterprise
design techniques
Beginning
optimization (MDO)
of Lifecycle
Manufacturing
Conceive
assembly
“process information”
Design
integration
create
creativity
architecting
trade studies
virtual
choose
SRR
- Mission
- Requirements
- Constraints
CDR
PDR
iterate
Implement
The System
iterate
real
Customer
Stakeholder
User
Architect
Designer
System Engineer
1
“turn
information
to matter”
The Environment:
Environment technological, economic, political, social, nature
Lifecycle Management (cont.)
The Environment:
Environment technological, economic, political, social, nature
testing
validation
verification
The Enterprise
Operate
service
test
Upgrade
Architect
Designer
System
Engineer
deploy
1
AR
The System
accept
control
usage
Customer
Stakeholder
System ID
User
behavior
prediction
real
End of
Lifecycle
monitor
virtual
monitor
Liquidate
EOL
control
usage
degrade
NASA Life-Cycle Phases
NASA Life
Cycle Phases
FORMULATION
Pre-Systems Acquisition
Pre-Phase A:
Concept
Studies
Project
Life Cycle
Phases
Project
Life Cycle
Gates &
Major Events
Agency
Reviews
Phase A:
Concept & Technology
Development
KDP B
KDP A
Preliminary
Project Plan
Draft Project
Requirements
2.
3.
4.
5.
6.
7.
Phase C:
Final Design &
Fabrication
KDP C
Phase D:
System Assembly,
Int & Test, Launch
KDP D
Phase E:
Operations
& Sustainment
KDP F
KDP E
End of Mission
Decommissioning
Phase F:
Closeout
Final Archival
of Data
Baseline
Project Plan7
ASM5
MCR
SRR SDR
(PNAR)
PDR
(NAR)
CDR /
PRR2
SIR
Re-enters appropriate life cycle phase if
modifications are needed between flights6
Robotic
Mission Project
Reviews1
1.
Operations
Launch
Re-flights
Launch
Readiness
Reviews
Supporting
Reviews
FOOTNOTES
Phase B:
Preliminary Design &
Technology Completion
IMPLEMENTATION
Systems Acquisition
FAD
ASP5
Human Space
Flight Project
Reviews1
Approval for
Implementation
16
MCR
SRR MDR4
(PNAR)
PDR
(NAR)
CDR /
PRR2
ORR
SAR
Inspections and
Refurbishment
FRR PLAR CERR3 End of
Flight
DR
PFAR
SIR
ORR
FRR PLAR
CERR3
DR
SMSR, LRR
(LV), FRR (LV)
Peer Reviews, Subsystem PDRs, Subsystem CDRs, and System Reviews
Flexibility is allowed in the timing, number, and content of reviews as long as the
equivalent information is provided at each KDP and the approach is fully
documented in the Project Plan. These reviews are conducted by the project for
the independent SRB. See Section 2.5 and Table 2-6.
PRR needed for multiple (≥4) system copies. Timing is notional.
CERRs are established at the discretion of Program Offices.
For robotic missions, the SRR and the MDR may be combined.
The ASP and ASM are Agency reviews, not life-cycle reviews.
Includes recertification, as required.
Project Plans are baselined at KDP C and are reviewed and updated as
required, to ensure project content, cost, and budget remain consistent.
ACRONYMS
ASP—Acquisition Strategy Planning Meeting
ASM—Acquisition Strategy Meeting
CDR—Critical Design Review
CERR—Critical Events Readiness Review
DR—Decommissioning Review
FAD—Formulation Authorization Document
FRR—Flight Readiness Review
KDP—Key Decision Point
LRR—Launch Readiness Review
MCR—Mission Concept Review
MDR—Mission Definition Review
NAR—Non-Advocate Review
ORR—Operational Readiness Review
PDR—Preliminary Design Review
PFAR—Post-Flight Assessment Review
PLAR—Post-Launch Assessment Review
PNAR—Preliminary Non-Advocate Review
PRR—Production Readiness Review
SAR—System Acceptance Review
SDR—System Definition Review
SIR—System Integration Review
SMSR—Safety and Mission Success Review
SRR—System Requirements Review
Lifecycle Modeling
7
ISO/IE 15288 Lifecycle Processes
• What should we model?
8
Optimal Lifecycle Design
• Traditionally, design has focused on performance
e.g. for aircraft design
optimal = minimum weight
• Increasingly, cost becomes important
• 85% of total lifecycle cost is locked in by the end of
preliminary design.
• But minimum weight ≠ minimum cost ≠ maximum value
• What is an appropriate value metric?
9
Design Example
•
•
•
We need to design a particular portion of the wing
Traditional approach: balance the aero & structural requirements,
minimize weight
We should consider cost: what about an option that is very cheap to
manufacture but performance is worse?
manufacturing cost?
aircraft demand?
aircraft price?
•
•
•
•
10
•
aerodynamics?
structural dynamics?
tooling?
environmental impact?
How do we trade performance and cost?
How much performance are we willing to give up for $100 saved?
What is the impact of the low-cost design on price and demand of
this aircraft?
What is the impact of this design decision on the other aircraft I
build?
What about market uncertainty?
Value Optimization Framework
Manufacturing
Tooling
Design
Operation
Aerodynamics
Structures
Weights
Mission
Stability & Control
Market factors
Fleet parameters
Competition
11
Cost
Module
Performance
Module
Revenue
Module
“Value” metric
“Optimal”
design
Challenges
•
•
•
•
•
•
12
Cost and revenue are difficult to model
– often models are based on empirical data
– how to predict for new designs
Uncertainty of market
Long program length
Time value of money
Valuing flexibility
Performance/financial groups even more uncoupled
than engineering disciplines
Cost Model
Cost
Module
Performance
Module
Revenue
Module
13
“Value” metric
Need to model the lifecycle
cost of the system.
Life cycle :
Design - Manufacture Operation - Disposal
Lifecycle cost :
Total cost of program over
life cycle
85% of Total LCC is locked
in by the end of preliminary
design.
Impact on LCC (%)
14
60
40
20
0
65%
Time
(From Roskam, Figure 2.3)
Disposal
Operation
and support
Manufacturing
and acquisition
80
Detailed design
100
Preliminary design,
system integration
Conceptual
design
Lifecycle Cost
95%
85%
15
Tooling
Other
Cost incurred one time only:
Engineering
- airframe design/analysis
- configuration control
- systems engineering
Tooling
- design of tools and fixtures
- fabrication of tools and fixtures
Other
- development support
- flight testing
Engineering
Non-Recurring Cost
Cost Estimating Methods
Basic techniques to develop Cost Models:
16
(1)
Detailed bottom-up estimating
- identify and specify lower level elements
- estimated cost of system is Σ of these
- time consuming, not appropriate early, accurate
(2)
Analogous Estimating
- look at similar item/system as a baseline
- adjust to account for different size and complexity
- can be applied at different levels
(3)
Parametric Estimating
- uses Cost Estimation Relationships (CER’s)
- needed to find theoretical first unit (TFU) cost
Cost Breakdown Structure (CBS)
Organizational Table that collects costs, covers:
- research, development, test and evaluation (RDT&E)
- production, including learning curve effects
- launch and deployment
- operations
Operations
- end-of-life (EOL) disposal
Production
RDT&E
Space
Mission
Architecture
Program
Level Costs
• Management
• Systems Eng
• Integration
17
Space
Segment
• Payload
• Spacecraft
• Software
• “Systems”
Launch
Segment
Ground
Segment
Operations
and Support
• Launch Vhc
• Launch Ops
• S/C-L/V
integration
• Facilities
• Equipment
• Software
• etc
• Personnel
• Training
• Maintenance
• Spares
Parametric Cost Models
Are most appropriate for trade studies:
Advantages:
• less time consuming than traditional bottom-up estimates
• more effective in performing cost trades
• more consistent estimates
• traceable to specific class of aerospace systems
Major Limitations:
• applicable only to parametric range of historical data
• lacking new technology factors, adjust CER to account for new technology
• composed of different mix of “things” in element to be costed
• usually not accurate enough for a proposal bid
18
Process for developing CER’s
Subsystem N
….
Subsystem B
Subsystem A
Cost/parametric data
Constant Year Costs
Computer
Software
Regression
Analysis
Preferred Form
$ = AW b
Step 1
Develop Database
File
19
Step 2
Apply Regression
Analysis
(Cost Model Assumption)
Subsystem N
….
Subsystem B
Subsystem A
$
Weight - kg
Key statistics: R2, ρ
Standard Error: RMS, σ
Step 3
Obtain CER’s and
Error Statistics
Adjustment to constant-year
dollars
It is critical that cost estimated be based on a
constant-year dollar bases. Reason: INFLATION
E.g. All costs are adjusted to FY92 (“Fiscal Year 1992”)
CY = R ⋅ CY − N
Past Years
Use actual inflation numbers
R = (1.040 )(1.037 )(1.034 ) = 1.115
123 1
424
3 123
FY 92
FY 93
Future Years
R = (1 + iRATE )
20
N
FY 94
Convert Oct-1991 cost
to Oct-1994 costs
Use forecasted inflation numbers
e.g. 3.1% yearly inflation in U.S.
$ 1M in FY 1980 corresponds to
$ 2.948M in FY 2005
Development Cost Model
• Cashflow profiles based on beta curve:
c(t ) = Kt α −1 (1 − t ) β −1
• Typical development time ~6 years
• Learning effects captured – span, cost
0.06
Support
0.05
normalized cost
Tool Fab
0.04
Tool Design
ME
0.02
(from Markish)
Engineering
0.01
0
normalized time
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53
21
22
Material
Support
Cost incurred per unit:
Labor
- fabrication
- assembly
- integration
Material to manufacture
- raw material
- purchased outside
production
- purchased equipment
Production support
- QA
- production tooling support
- engineering support
Labor
Recurring Cost
Learning Curve
As more units are made, the recurring cost per
unit decreases.
This is the learning curve effect.
e.g. Fabrication is done more quickly, less
material is wasted.
Yx = Y0 x
n
Yx = number of hours to produce unit x
n = log b/log 2
b = learning curve factor (~80-100%)
23
Learning Curve
1
Cost of unit
0.8
0.6
0.4
b=0.9
0.55
0.2
Every time
production
doubles,
cost is
reduced by
a factor of
0.9
0
0
10
20
30
40
50
Unit number
Typical LC slopes: Fab 90%, Assembly 75%, Material 98%
24
Airplane Related Operating Costs
CAPITAL COSTS:
Financing
Insurance
Depreciation
Capital
Costs
CAROC
40%
60%
CASH AIRPLANE RELATED
OPERATING COSTS:
Crew
Fuel
Maintenance
Landing
Ground Handling
GPE Depreciation
GPE Maintenance
Control & Communications
CAROC is only 60% - ownership costs are significant!
25
Value Metric
Need to provide a
quantitative metric that
incorporates cost,
performance and
revenue information.
Cost
Module
Performance
Module
Revenue
Module
26
“Value” metric
In optimization, need to
be especially carefully
about what metric we
choose...
Value Metrics
Traditional Metrics
Augmented Metrics
performance
weight
speed
cost
revenue
profit
quietness
emissions
commonality
...
The definition of value will vary depending on your system
and your role as a stakeholder, but we must define a
quantifiable metric.
27
Net Present Value (NPV)
• Measure of present value of various cash flows in different
periods in the future
• Cash flow in any given period discounted by the value of a
dollar today at that point in the future
– “Time is money”
– A dollar tomorrow is worth less today since if properly
invested, a dollar today would be worth more tomorrow
• Rate at which future cash flows are discounted is
determined by the “discount rate” or “hurdle rate”
– Discount rate is equal to the amount of interest the
investor could earn in a single time period (usually a
year) if s/he were to invest in a “safer” investment
28
Discounted Cash Flow (DCF)
• Forecast the cash flows, C0, C1, ..., CT of the project
over its economic life
– Treat investments as negative cash flow
• Determine the appropriate opportunity cost of capital
(i.e. determine the discount rate r)
• Use opportunity cost of capital to discount the future
cash flow of the project
• Sum the discounted cash flows to get the net present
value (NPV)
C1
C2
CT
+
+K+
NPV = C0 +
2
T
1+ r (1+ r )
(1+ r )
29
DCF example
Period
Discount Factor
Cash Flow
Present Value
0
1
-150,000
-150,000
1
0.935
-100,000
-93,500
2
0.873
+300000
+261,000
Discount rate = 7%
30
NPV =
$18,400
Net Present Value (NPV)
T
Ct
NPV = ∑
t
t =0 (1 + r )
1500
500
29
27
25
23
21
19
17
15
13
11
9
7
5
3
0
1
Cashflow, Pt [$]
1000
Cashflow
DCF (r=12%)
-500
-1000
-1500
Program Time, t [yrs]
31
© Massachusetts Institute of Technology - Prof. de Weck and Prof. Willcox
Engineering Systems Division and Dept. of Aeronautics and Astronautics
Return on Investment (ROI)
• Return of an action divided by the cost of that
action
revenue − cost
ROI =
cost
• Need to decide whether to use actual or
discounted cashflows
32
© Massachusetts Institute of Technology - Prof. de Weck and Prof. Willcox
Engineering Systems Division and Dept. of Aeronautics and Astronautics
Traditional Design Optimization
¾ Objective function:
usually minimum
weight
¾ Design vector:
attributes of design,
e.g. planform
geometry
¾ Performance model:
contains several
engineering
disciplines
33
Performance
Model
Objective
function J(x)
Design
vector x
Optimizer
Coupled MDO Framework
¾Objective function: value
metric, e.g. NPV
¾Simulation model:
performance and financial
¾Stochastic element
Performance
Model
Market
Cost
Cost
Valuation
Revenue
Price,
Demand
Design
vector x
Optimizer
34
σ, α
Objective
function J(x)
Summary
• Lifecycle Management
– Operations phase is often the longest and most expensive
– Design for maintainability, upgrades, evolution …
• Lifecycle Modeling
– Cost = Non-recurring + Recurring, Fixed + Variable
– Revenue, Value
– Others … e.g. energy consumption, carbon footprint …
• Take 16.888 Multidisciplinary System Design
Optimization in Spring 2010 if you want more !
• Online final exam will be posted this weekend by Dec 6,
2009 at the latest – 4 days to respond (open book)
• Friday, Dec 11 – social event (LEGO Mind Storms)
35
Thank you!
TA: Maj. Jeremy Agte … could not have done it without you !
Happy Holidays !
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.