product development processesdspace.mit.edu/bitstream/handle/1721.1/80702/esd... ·...
TRANSCRIPT
+
- Product Development Processes
ESD.36J System & Project Management
System Project Managementand
Instructor(s)
Prof. Olivier de Weck
Lecture 11
10/9/2003
+
- Introduction
� Project Preparation and Planning � Importance of Upfront Preparation
� Contents of Program Management Plan
� Product Development Processes (PDP) � Waterfall Model � Spiral Model
Acknowledgments: Prof. Ed Crawley Prof. Steve Eppinger Dr. Joyce Warmkessel Dr. Darian Unger
10/9/2003 - ESD.36J SPM 2
+
-ESD.36 Framework
Project Project Project Preparation Planning Execution
Doing the right job
Project Adaptation
Doing the job right
10/9/2003 - ESD.36J SPM 3
+
- Preparation versus Planning
� “Preparation” address the who, what and why issues � Who are the major stakeholders? � Why we are doing the project? � What do the stakeholders expect in exchange for
their participation (“value”)? � Time frame for “Preparation” is prior to
project initiation � “Planning” is the “how” the project will be
conducted
10/9/2003 - ESD.36J SPM 4
+
-
Management Leverage is Greatest Early
10/9/2003 - ESD.36J SPM 5
I
+
- Preparation Actions
e
Res
Identify program
processes
Dd eOS ten et rI S raN t gd mtt i aoS fe reeed y U
n in u ut E na it rces
c z enik s tf aunde y tri-ho
iqU oeu nsl ede rrs
Planning
no p eiss
Mi
t
c o
St c t
tse nec
j yuB/ekaM
j fyoce o r sserddA
ariir tPy
secruoseR
orPpo
j nedI
Pe
puen
snoCt
escif an tt
n fneme
m i
n g
L iredeeveDl
e r
tmi
t nIe T&
iat
D ItS
10/9/2003 - ESD.36J SPM 6
i
+
- Planning Actions
DDeveleP teo A rp p muS o pl p Pr
iD t a r nr t oe u e Re Ov aceR t cou Sl h igo re sr t rgq p P r k Mae aa uu m Sn cir r i tze o uam d nr ct e W
hu aie o ec gn n dt et a us mBl leS en Executable
Program Plan
t
esiar nla ordn P n t
nt &m f c en
anm eou ioes csm b ti ite geassm
et mc s rai ty hl i ni e s or enea acb&
M oe Tn Ma
mmus i rf sse rn eac pPlponn p A pf
e Co ioe es ne es AD inetd r rat i
m uf i ole rn efu Dan D nta sfi orPe efD Mn CoI
10/9/2003 - ESD.36J SPM 7
+
- Program Management Plan Topics
� Summary -program description and major objectives
� Assumptions & constraints
� Deliverables � Work breakdown
structure � Major milestones
with associated exit criteria
� Staffing& training plan
� Spending/budget profile
� Project organization � External interfaces � Program plans: list
with summary description
� Product documentation tree
10/9/2003 - ESD.36J SPM 8
+
- Other Lower Level Plans � Control plans
� Schedule/budget � Requirements � Configuration � Technical Parameters � Interface
� Risk Management � Subcontract
Management � Contract
Management
� Reviews and Audits � Quality Assurance � Safety � Reliability � Problem resolution � Information
Management
10/9/2003 - ESD.36J SPM 9
+
-
Four Basic Tensions (Trade-offs) in Product/System Development
Performance
Schedule Risk
Cost Ref: Maier/Rechtin
One of the main jobs of the program manager (and architect) is to identify the principle tensions and resolve them
- ESD.36J SPM 10
+
- Factors that Influence the Program Plan
Development Program StructureApproach • New Product • Single organization
• Update • Prime with • Platform Product subs/suppliers
• Platform • Federation of development associates
Product Strategy Execution • New technology • Does “culture” or customer
• Significant reuse require specific execution
• Integration of existing approach (V, spiral etc)
“subsystems” • Platform based
10/9/2003 - ESD.36J SPM 11
+
- Characteristics of PDP Always involves at least:• Marketing core • Design functions
Challenges of the PDP • Manufacturing
1. Tradeoffs2. Dynamics3. Details “A PDP is the unique sequence4. Time Pressure of steps or activities, which an5. Economics enterprise employs to conceive,
design, and commercialize a product”
10/9/2003 - ESD.36J SPM
Ulrich and Eppinger 12
- ESD.36J SPM10/9/2003 13
+
-
1
Beginningof Lifecycle
- Mission- Requirements- Constraints Customer
StakeholderUser
ArchitectDesignerSystem Engineer
ConceiveDesign
Implement
“process information”
“turninformation
to matter”
SRR
PDR CDR
iterate
iterate
The EnvironmentThe Environment: technological, economic, political, social, nature
The EnterpriseThe Enterprise
The SystemThe System
creativityarchitectingtrade studies
modeling simulationexperiments
design techniquesoptimization (MDO)
virtual
real
Manufacturingassembly
integration
choose
create
Conceive, Design, Implement
- ESD.36J SPM10/9/2003 14
+
-
End of Lifecycle
1
The SystemThe Systemreal
AR
virtual
CustomerStakeholderUser
EOL
testdeploy
service
monitor
OperateUpgrade
Liquidate
degrade
ArchitectDesignerSystem Engineer
accept
controlusage
testing validation
verificationThe EnterpriseThe Enterprise
monitor
controlusage
The EnvironmentThe Environment: technological, economic, political, social, nature
System IDbehavior
prediction
Operate, Upgrade, Liquidate
+
- Generic PDP Process
Need Assessment
Define Requirements
Concept Evaluation
Internal and external needs.
Future needs
DevelopDesign Test Implement
Trade-off Studies
Materials, Geometries
Tooling, Detailed Specifications.
Manufacturing Techniques
Formulate and Translate Customer Wants
Evaluate and Needs into Engineering
Concept Requirements.
Alternatives Top down System to Components.
System to System Requirements
Prototypes, Verify, Validate Design
Build and and Qualify Completion .
Integrate Transition to
Subsystem Test Operations.
Lessons Learned
10/9/2003 - ESD.36J SPM 15
Roles in PDP*+
-
*Ulrich,K. and Eppinger,S. “Product design and development”, McGraw-Hill, 1995, p15.
10/9/2003 - ESD.36J SPM 16
+
- Key Differences in PDP
� Number of phases (often a superficial difference) � Phase exit criteria (and degree of formality) � Requirement “enforcement” � Reviews � Prototyping � Testing and Validation � Timing for committing capital � Degree of “customer” selling and interference � Degree of explicit/implicit iteration (waterfall or not) � Timing and degree of supplier involvement
10/9/2003 - ESD.36J SPM 17
+
-
PDP In the SMALL v. In the LARGE
Product:
Org
Product
Process
SIMPLE Countable interfaces
Process: Parts identifiable
COMPLEX Very many interfaces Parts abstracted
STATICGoals & resources
EVOLVING Goals & resources
changingOrganization: constant
LARGE SMALL Big team
Team at a table
Other Factors: clean slate vs legacy components; single product vs platform; collocated vs distributed team; team in enterprise vs supplier involvement
10/9/2003 - ESD.36J SPM 18
+
-Example: Global Telecommunications Services Development Process
� Multi-national telecommunications company � “Document-based” development environment � 100+ documents involved in complex new IP product � 15+ internal organizations
� technical, billing, product, service, etc.
� Redundant information and duplicated efforts � Too many projects � Frustrating and long development cycle
We have the process on paper – but we don’t live it.
Ref: Eppinger 10/9/2003 - ESD.36J SPM 19
+
-
Telecommunications Development: The Planned Process
Concept Approval CONCEPT
SPECIFICATIONS
Network Plan
Req’ts
Eng’g Specs
Specification Approval
OPERATE
�Four-Phase Development Process �150 Documents in DSM Model
Business Req’ts
Ops.
Implem. & Billing
Cust. Service
Eng’g DETAILED DESIGN
Launch Approval
PLANNING AND
System
AND BUILD
10/9/2003 - ESD.36J SPM 20
+
-
Telecommunications Development: The Actual Process
Concept Approval CONCEPT
SPECIFICATIONS
Network Plan
Req’ts
Eng’g Specs
OPERATE
Business Req’ts
Ops.
Implem. & Billing
Cust. Service
Eng’g DETAILED DESIGN
Launch Approval
Critical Problem:
PLANNING AND
System
AND BUILD
Ongoing Iterations after Specifications Phase
10/9/2003 - ESD.36J SPM 21
+
-
Telecommunications Development: An Improved Process
�
�
�
� Financial analysis
Concept ApprovalCONCEPT
SPECIFICATIONS
Network Plan
Req’ts
Eng’g Specs
Specification Approval
OPERATE
Business Req’ts
Ops.
Implem. & Billing
Cust. Service
Eng’g DETAILED DESIGN
Launch Approval
Solution: �
�
�
� Enables Parallel Design Efforts
!
Gate is Critical: Frozen product requirements
Project plan for detailed design
PLANNING AND
System
AND BUILD
Planned Iterations in Specs Phase X-Functional Resources in Specs Formal Spec Approval Gate
Specification Approval
Frozen technical specifications
10/9/2003 - ESD.36J SPM 22
+
-Phased vs. Spiral PD Processes
Phased, Staged, or Waterfall PD Process(dominant for over 30 years)
LaunchDefinition Level Design
Detail Design and Test
Product Planning
Product Product System- Integrate
Spiral PD Process(primarily used in software development)
Launch
Define, Design, Build, Test, Integrate
Define, Design, Build, Test, Integrate
Define, Design, Build, Test, Integrate
Product Planning
Product
Process Design Questions:� How many spirals should be planned?
� Which phases should be in each spiral?
�
10/9/2003 - ESD.36J SPM
When to conduct gate reviews? 23
Stage Gate PD Process +
-
-10/9/2003 24
Release
Planning
Design
Design
Detailed Design
Integration & Test
Reviews
Iterations (planned)
Cross-Phase Iterations
(unplanned)
Refs: Robert Cooper, Winning at New Products 3rd ed., 2001.
ESD.36J SPM
Concept
System-Level
Within-Phase
Spiral PD Process+
-
10/9/2003 - ESD.36J SPM 25
Risk Analysis
Prototype 1
Concept of operation
Requirementsvalidation
Design validation andverification
Release
Acceptance test
Integrationand test
Unittest
Code
Software
Softwarerequire-ments
Requirementsplan, lifecycle plan
Development plan
Integration and test plan
Plan the nextiteration
PartitionReview
Commit to an approach for the
next iteration
Determineobjectives, alternatives,
and constraints
Identify and
resolve risks
Evaluatealternatives
Develop the deliverables for theiteration and verify that
they are correct
Cumulative cost
Detailed design
benchmarksmodels,
Simulations,
Prototype 2Prototype 3
Operational prototype
Risk Analysis
Risk Analysis
Risk Analysis
START
product design
Steve McConnell, Rapid Development, 1996, pp. 136-39.
Barry Boehm, A Spiral Model of Software Development and Enhancement, 1988.
Adapted from:
+
- Spiral PD Process (Simplified)
Planning
Detailed Design
Integration & Test
System-Level Design
Design
(Cumulative Effort) Release
Concept
Reviews
Cost
10/9/2003 - ESD.36J SPM 26
Spiral PD Process+
-
Design
Design
Integration & Test
Release
Reviews
Planning
Concept
System-Level Design
Detailed
Time
10/9/2003 - ESD.36J SPM 27
Concept
Requirementanalysis
Medium High Priority: Detailed design, test
High Priority: Detailed design, test
Medium Priority: Detailed design, test
Medium Low Priority: Detailed design, test
Low Priority: Detailed design, test
+
-
Design-to-Schedule, Design-to-Budget PD Processes
Release
s
Planning
Design
Sys-Level Design
Detailed design, test
Detailed design, test Run out of time or budget here
Concept
Medium-High Priority:
High Priority: Detailed design, test
Medium Priority:
Medium-Low Priority: Detailed design, test
Low Priority: Detailed design, test
10/9/2003 - ESD.36J SPM 28
+
-Other PD Processes
� Evolutionary delivery � Evolutionary prototyping � Extreme programming (software only) � Modified stage gates � Set-based concurrent engineering � Short-term solutions, i.e. acquisitions
Ref: Steve McConnell, Rapid Development, 1996, pp. 136-39.
10/9/2003 - ESD.36J SPM 29
(Hybrid) Case Study: Xerox +
-
� Hardware (DC 460-ST) and software (Endeavor) � Large company and steady platform (variant) products � Iterations
� TTM process technically a stage gate but w/some important exceptions � Software subprocess
� Internal experiments and waivers
� Risk – emphasis on controlling schedule
10/9/2003 - ESD.36J SPM 30
+
-Xerox Time-to-Market (TTM) Process
Define market attack plan & technology
Define product & deliver
Design Demonstrate Delight
technology product product
Deliver product customers
Software Development Sub-Process
10/9/2003 - ESD.36J SPM 31
+
-
Product Development Process Design
� There are many possible PD processes to choose from. � Spectrum from staged to spiral processes
� PD processes should be designed to address project risks by � Planning iterations in the process
� Can be within or across phases
� Scheduling reviews to control the process � Cross-functional review � Generally reviews end phases (no backtracking)
10/9/2003 - ESD.36J SPM 32
+
- Proposed PDP Design Procedure
1. Identify and prioritize the project risks. 2. Assign each risks to a specific phase. 3. Plan the necessary iteration cycles within
each stage to address the assigned risks. 4. Schedule reviews at the completion of
each phase. Review ReviewReview Phase Phase Phase
Risks Risks Risks
10/9/2003 - ESD.36J SPM 33
+
- Key similarities between PDPs
� Attempt to meet customer needs � Discrete phases of development � Iteration in some form � Existence of reviews or gates � Attempt to control development risks
10/9/2003 - ESD.36J SPM 34
+
- Key differences between PDPs
� Names and number of development phases � Type of iteration � Type of review � Customer interface points � Product complexity � Risks addressed
10/9/2003 - ESD.36J SPM 35
+
- Identification of key risks
Risk Profile
Category of risk
Technical/ Performance
Market/ User
Budget
Schedule
1 2 3 4 5 Low High (Coord./Variation) (Unk-unks/chaotic)
Level of risk
10/9/2003 - ESD.36J SPM 36
+
-Iteration parameters �
�
�
10/9/2003
Breadth of iterations1 2 3
Narrow Comprehensive(Within 1phase) (Across 3 or more phases)
Number of interphase loops
0 1 2 3 4
planning
Concept design
System level design
Detailed design
System testing
release
Product
Product
None 4 or more(No iteration) (Multiple iterations)
Level of planning1 2 3 4 5
None Planned but Planned (Unexpected)
- ESD.36J SPM & schedulednot scheduled
37
+
- Review/Gate parameters
Format of stage
gate
� Rigidity
Final standard Phase check ( igi
� Frequency
Frequent Sporadic (After each phase) (After 3 or more phases)
1 2 3 4 5
1 2 3 4 5 Outputs
Review /OK?
Next stage
Deliverables
Yes
NoRigid criteria) (Less r d criteria)
Action stage
10/9/2003 - ESD.36J SPM 38
+
-Parameterized theoretical PDPs
Parameter Waterfall/ Design to Evolutionary Spiral Stage-gate Sched/Budget prototyping
Start End
Iterations
Breadth 1 2 1 3 3
# of interphase loops
0 Unspecified 0 Unspecified
Unspecified, but many
Planning 2 3 2 3 5
Rigidity 1 2 3 3 5 Review gates
Frequency 1 2 2 1 Unspecified
Risk Profile of Manages tech Manages Manages Manages key risks risk well sched./budget market risk well market risk
risk well well
10/9/2003 - ESD.36J SPM 39
IDeSat. Rad.
s
Ref: D. Unger+
- Parameterized actual PDPs
Market risk dominates
5
4
4
3
2
Schedule is king
1 2-3
1 4
3 4
0 3
1 3
DC SW
3 3-43Planning
Little market ambi
ProfileRisk
gates
Iterations
1 21
3 41Rigidity
1 2 phase loops
1 22
ITTParameter
guity Unknown
Review
Frequency
Unknown # of inter
Breadth
Sikor ky
10/9/2003 - ESD.36J SPM 40
+
- “Tradeoff” in PDP design
� Predictability � Early spec definition
� Narrow iteration
� Frequent, rigid reviews
� Reduces technical risk
• Flexibility – Market tests to
ensure viability – Broad iterations – Loose, multi-phase
reviews – Reduces market risk
10/9/2003 - ESD.36J SPM 41
+
- IDEO Product Case
� What are the characteristics of the IDEO product development process? � …
10/9/2003 - ESD.36J SPM 42
+
- Conclusions
� Preparation and Planning are important � PDPs are “project plan templates” � Phase-Gate versus Spiral continuum � Hybrid forms in real company contexts � PDP design driven by dominant risks
� Technical, market, schedule, budget
� Key distinguishers/PDP design factors � Iterations: breadth, number, planned/unplanned � Gates: rigidity (rigor), frequency
10/9/2003 - ESD.36J SPM 43