Title: Project Management Jan 27, 2006
1Project ManagementJan 27, 2006
- James R. Matt
- Technical Fellow
- General Motors Corp
2I think there is a world market for maybe five
computers.
Conventional Wisdom A Rogues Gallery
-- Thomas Watson, chairman of IBM, 1943
There is no reason anyone would want a computer
in their home.
Ken Olson, President, Chairman and Founder of
Digital Equipment Corp., 1977
640K ought to be enough for anybody.
Bill Gates, 1981
What do 13 people in Seattle know that we dont?
Ross Perot when presented with a proposal for
EDS to acquire Microsoft, 1980
3Key Thoughts Simple Tools
4Coarse to FineProduct Development
Proof of Concept
Production Readiness
Manufacturing Production
Select
Technical Solution Development
Organize Prioritize
Needs Ideas
Customer and Market Feedback (data from
Marketing, Sales, Quality, Benchmarking,
Customers)
5Spider Chart Performance Targets and Key Wins
Smaller Size
Faster Speed
Durability Life
Key Wins
Todays Product or Situation
Best in Class Competition
Targets for new Design
6Trade-off study matrix
Design Option 1
Design Option 2
Design Option 3
Example Criteria
Performance Cost Mass Quality Volume /
Size Risk Durability
Key Much Better Somewhat
Better 0 No Improvement - Worse
- - Much Worse
Summation
7Trade-off study matrix
Design Option 1
Design Option 2
Design Option 3
Example Criteria
Performance Cost Mass Quality Volume /
Size Risk Durability
- -
Key Much Better Somewhat
Better 0 No Improvement - Worse
- - Much Worse
-
0
0
0
0
-
0
- -
-
1
5
3
Summation
8Engineering Project Management
9DFSS
IDDOV
Product Development
Determine Needs, Select Projects. Assign Teams
Phase 00
Phase 00
Product Development
Technology
Planning
Define
Optimize
Requirements,
Generate
Detail
Key Interfaces, Constraints
Concepts
Concept
Verify
1
2
5
4
3
Select Approve
Approve Project
Requirements
Approve Detail Concept
Approve Concept
Plan, Team, and
Design Concept
Agreement
Deliverables
Verification
Design for Six Sigma IDDOV
I
-
Identify
D
1
Define Requirements
D
Design Concept
2
O
-
Optimize
V
-
Verify
10Project Tasks and Gate Reviews
Phase 00
Phase 00
Product Development
Define
Optimize
Determine Needs, Select Projects. Assign Teams
Requirements,
Generate
Detail
Technology
Key Interfaces, Constraints
Planning
Concepts
Concept
Verify
Define Requirements
Concept Generation
Robust Assessment
Optimize Validate
Define Initial Project
Key Interfaces Constraints
Design Selection
Purchasing
Tasks
Tasks
Tasks
Tasks
Determine Perf Manf
Tasks
Develop Design
Develop purchasing info
Generate Assess
Requirements
Establish Project Plan
Finalize Development Test Plan
Alternative concepts
Conduct Robust assessment
Define Business Targets
Obtain Lessons Learned
Optimize Concepts
Perform Evaluations
Construct, Build, and Test
Define Needs
Prototype
Draft Initial Specifications
Concept Tradeoff Study
Conduct Peer Review
Verify Hardware, Software
Develop Commercial
Define Interfaces, Constraints
to Technical Requirements
Approach
Refine Specifications
Gather Information to
Update all Business,
Robust
Eng Plan (DOE)
Define Project Plan
determine Requirements
Technical, and Project
Update all Business
Resources Required
Documents
Technical Documents
Conduct Gate Review 1
Conduct Concept Review and Approve Design Concept
Review with
Approve Details Concept Make Purchasing Decisions
Conduct Final Design Review
key Stakeholders
1
2
4
3
5
Select Approve
Approve Concept
Approve Project
Approve Detail Concept
Requirements
Design Concept
Plan, Team, and
Agreement
Verification
Deliverables
11Five Objectives of Every Project Gate Review
- Explain the Benefits of the Technical Solution or
Technology - Show the Technical Feasibility of approach and
solution - Show the Balance of Performance to Business
Imperatives - 4) Explain the Risks
- a) Business.
- b) Application Timing.
- c) Technical.
- 5) Explain the Expected Engineering Expenses
Costs.
12Risk
13Risk
- Types of Risk
- (things gone wrong, or critical items not going
right) - Business
- Timing
- Technical
- Quantify Relative Risk
- Risk f(L,M) (Likelihood) X (Magnitude)
- Method
- Brainstorm Potential Problems
- Define Likelihood of Occurrence (1 -10 scale)
- Define Magnitude should problem occur (1-10
scale) - (Risk Priority Number) RPN (Likelihood) X
(Magnitude) - Rank order into a bar chart (Pareto Diagram)
- Define Countermeasures
14Pareto Diagram
Focus on the high RPN Risk items and put in place
Countermeasures Must Insure -- Bang for the
Buck (limited resources cause the need to Focus)
Higher
RPN (Risk Priority Number)
Lower
Specific Potential Problems
15Modularity Standard Parts as Risk Reduction
16Component Design Build Approach
Transfer Function
Intended Output
Input
Risk Output not Design Intent
17Component Design Build Approach
Transfer Function
Input
Intended Output
Transfer Function
Input
Risk Unintended Interactions
18Independent Component Design Build Approach
19Component Design Build Approach
20Modular Build Approach
Sub-assembled Pre-Tested Modules
21Modular Build Approachwith Standard Parts
Using Standard Parts
22ExampleHeadlamp Turn SignalMultifunction
Electrical Switch
235 Phase Problem Resolution
- Problem Definition
- Containment, Immediate Corrective Action
- Root Cause
- Irreversible Corrective Action
- Verification
24ExampleMultifunction Headlamp Switch
Functions
Headlamp On / Off
Headlamp High / Low
Electrical Control
Customer
Turn Signal
Component test plan was designed to test each
function as independent variables All part
passed the lab test without incident
Cruise Control Set, On, Off
25ExampleMultifunction Headlamp Switch
Functions
Headlamp On / Off
Headlamp High / Low
Customer
Electrical Control
Turn Signal
However Mechanical Interaction Inside
the Multifunction Switch caused the headlamp
contact carrier to slightly rock when the Turn
Signal was used.
Cruise Control Set, On, Off
This caused a voltage spike and high resistance
path and heat in the switch.
26ExampleMultifunction Headlamp Switch
DOE factors for test matrix 1) Type of Lubricant
in the switch 2) Contact Material 3) Contact
Plating 4) Spring Pressure
The optimal combination was found and the design
was quickly changed. No field issues were found
27ExampleUpper Strut MountDesign For Six Sigma
Example
28(No Transcript)
29Define the Basic Functions
- Brainstorm a list of Basic Functions the Product
must provide. What does this thing need to do? - Use Verbs
- React
- Position
- Isolate
- Filter
- Rotate
- Limit
- Amplify
- etc
30Consider Classis Failure Mechanism that Cause
Failure Modes
- Creep
- Fracture
- Yield
- Physio-Chemical Instability
- Dimensional Incompatibility
- Contamination
- Vibration and Mechanical Shock
- Environmental
- Wear
31Consider Classis Failure Mechanism that Cause
Failure Modes
- Creep (relaxation and flow over time, plastic
movement, often accelerated with heat or high
loads) - Fracture (brittle failure due to sudden physical
overload, cracking) - Yield (Tensile or bending failure, permanent
deformation) - Physio-Chemical Instability (Chemical change in
material properties, Corrosion, UV instability,
chemical attack of solvent or lubricants, heat
aging of rubber) - Dimensional Incompatibility (Stack up of
tolerances, mis-positioning, flexing of base or
bracket, too big, too small) - Contamination (dirt, grit, dust, mixed materials)
- Vibration and Mechanical Shock (mechanical or
electrical high frequency, surge, sudden
overload) - Environmental (hot, cold, humid, submersion)
- Wear (repeated cyclic load causing material
removal)
32Matrix Functions vs Failure Mechanisms Evaluate
Risk due to Sensitivities
Failure Mechanisms
React Position Isolate Filter Rotate Limit A
mplify
Functions
33Matrix Functions vs Failure Mechanisms Evaluate
Risk due to Sensitivities
React Position Isolate Filter Rotate Limit A
mplify
High
Med
High
High
Med
High
34Matrix Assessment Knowledge Gathering
35DFSS - Front Upper Strut Mount
Position Stays in place over life Tolerant of
misaligned mating parts Isolate Quiet over
life Noise Transmission is good React
Loads Rate Curve is within Bandwidth Structure
handles load with out excessive Damage
(Failure Mechanisms cause Failure
Modes) Environment Contamination Yield Fatigue Fra
cture Vibration / Mechanical Shock Wear Electrical
/ Software Compatibility Physio-Chemical
Instability Creep Dimensional Incompatibility
36The Forward Pass
Likelihood of compromise of a Primary
Function due to a Sensitivity to Failure
Mechanism. Strong Likelihood Neutral
0 Not Sensitive
37The Reverse Pass learn for what has happened
before
Functions Position, Isolate, React Loads
Anticipated or Historic Problems
38Chart the probability of each Failure Mechanism
contributing to historic problems and loss of
desired Functions
Failure Mechanisms Environment Contamination Yiel
d Fracture Vibration / Mechanical
Shock Wear Electrical / Software Physio-Chemical
Instability Creep Dimensional Incompatibility
In this case, Contamination, Yield, Wear, and
Dimensional Incompatibility are the high
Occurrence Failure Mechanism that are anticipated
as the dominate Noise factors.
39Now think in terms of the design Components (and
for competing design options)
D
E
C
F
B
G
A
H
A Inner Metal B Primary/Shear Isolator C
Upper Rate Washer D Reaction Washer E
Reaction Isolator F Outer/Compression
Isolator G Main Stamping H Lower Rate Washer
40Top Mount Bearing
41- DOE
- Partial Factorial Matrix experiment
- Expert Knowledge, seek help in creation
- Garbage in Garbage out (usually due to bad
assumptions) - Want a simple lab fixture(s) to run a fast DOE to
understand design sensitivity to Noise factors. - Run carefully Created samples to test for
interactions
42Basic Principles Friction, transient loads,
moments, unforeseen interactions, manufacturing
processes capabilities, marginally stable
systems, static electricity, grounding Wrong
Assumptions (independence of functions in switch
example)Lack of parts available on
timeComponents cost more than estimatedStack
up of tolerances reality is not design
nominalMurphys Law, Chaos Theory, probability
and statistical theory at workFalse or
unachievable accuracy
Causes of Problems
43Idea GenerationDo not jump to conclusions or
select your mainstream concept too quickly.
Brainstorm, talk to experts, consider viability
of alternatives, carefully select your mainstream
direction based on Tradeoffs.ExecutionSet up a
good project plan with Milestones and Gate
ReviewsClearly defined Deliverables Set up a
budget (with a 10 contingency) and obtain good
cost estimates and availability of materials,
manpower, and facilitiesFront Load your efforts
get off to a good startAssign tasks to team
members based on skill sets and personal
preferences. Monitor progress, keep things
visible.Design and Build Modules where
practical.Use Standard parts and commonly
available materialsAllow time for
experimentation and Slack time for things to go
wrong in build test.Do your homework
Helpful Ideas Thoughts
44Closing Thoughts Good Luck Get Started Have
Fun Expect Things to go Wrong