lecture 13: project dynamics case studies
TRANSCRIPT
ESD.36 System Project Management
Instructor(s)
+
- Project Dynamics Applications and Cases
Dr. James Lyneis
Lecture 13
November 1, 2012
+
- Today’s Agenda
Managing Changes
• Case Example – Dispute Resolution • Case Example – Change Management
Project-to-Project Learning
• Case Example – Risk Management • Case Example – Bidding and Management
Broader issues – project portfolios, priorities and
market interactions
2
+
- SD Qualitative Insights – 3
4. Attempts to achieve an infeasible plan via project control actions lead to “vicious circle” side effects which increase project cost and duration.
On complex projects, these costs usually exceed the “direct” costs of infeasibility
5. Project “changes,” and risks which materialize, are fundamentally the same as an infeasible plan. (Lecture 13)
3
+
- Using an SD Model in Practice
1. Set up model to represent project plan (the plan should reflect normal amounts of rework)
2. Specify exogenous inputs for all changes to the plan as they occur(ed) (external and internal, including new policies or processes)
3. Refine parameter and change estimates via model calibration (usually after project completion)
4
+
-
Original Work to Do Work Done
Undiscovered Rework
Progressx
Quality
Time to Discover Rework
Rework to Do Rework
Discovery
- +
Productivity
Effort Applied
Rew ork
Progress Rework
Generation +
+ +
Scope Growth
Changes
Direct Changes to Project Plan
Project Time
Project Time
Project Time
1
Delayed Availability of Staff
Project Time
1
Late Vendor Info; Extra Mtgs
Project Time
1
Uncertain Specs & Reqmnts.
Typically many small changes, rather than one large change.
Fraction Correct and Complete
5
+
-
Project Behavior: 50 – 100 % Overrun
Time
Project Staffing
Typical Plan
Actual Results
6
Client (?) Expectation with changes
+
-
Staff100
75
50
25
00 6 12 18 24 30 36 42 48 54 60
Time (Month)
Peo
ple
Staff : SD5 ChangesStaff : SD5 Feasible Plan1
Work Done1,000
750
500
250
00 6 12 18 24 30 36 42 48 54 60
Time (Month)
Task
s
Work Done : SD5 ChangesWork Done : SD5 Feasible Plan1
Example: Changes Obsolete Work
Work Done Staff Level
“Changes” Plan “Changes”
Plan
20% @ Month 15
7% Tasks Redone 28% Cost Increase! 4X Multiplier
7
+
-
Fraction Correct and Complete1
0.85
0.7
0.55
0.40 6 12 18 24 30 36 42 48 54 60
Time (Month)
Frac
tion
Fraction Correct and Complete : SD5 ChangesFraction Correct and Complete : SD5 Feasible Plan1
Productivity1
0.9
0.8
0.7
0.60 6 12 18 24 30 36 42 48 54 60
Time (Month)
Task
/(Per
son*
Mon
th)
Productivity : SD5 ChangesProductivity : SD5 Feasible Plan1
Because of Lower Productivity and Fraction Correct
Productivity Fraction Correct
“Changes”
Plan
“Changes”
Plan
8
+
- Data on “Multiplier"
1 0 0
9 0
8 0
7 0
6 0
5 0
4 0
3 0
2 0
1 0
1 2 3 4 5 6 7 8 9 1 0
P e rc e n ta g e
o f
P ro je c ts
R a t io o f T o ta l C o st to D ire c t C o st
S o u rc e : P u g h -R o b e rts A sso c ia te s
D e la y a n d D is ru p tio n R a tio s
Costs snowball, increasing non-linearly with:
Cumulative magnitude of changes;
Duration and lateness in project
Data: Ratio of total cost increase to direct cost increase 3-5X
9
+
- Consequences …
If contractor/client relationship
Disputes
If internal development
Risks materialize
10
+
- Today’s Agenda
Managing Changes
• Case Example – Dispute Resolution • Case Example – Change Management
Project-to-Project Learning
• Case Example – Risk Management • Case Example – Bidding and Management
Broader issues – project priorities and market
interactions
11
+
-
How SD in Project Management got started: Ingalls Shipbuilding vs. US Navy
Ingalls Shipbuilding won contract for 9 LHA’s and 30 DD963 destroyers in 1969 and 1970.
Firm fixed price contract structure
$500 million cost overrun on the programs (Ingalls and Navy agreed on $150 direct cost – the rest?) Navy – bad management
Ingalls – D&D
Ingalls sues Navy claiming design changes caused delay and disruption
12
+
-
Planned Program
Conditions
Company Actions
Customer-Responsible Events & Conditions
Real Program
TIME
Total Program Labor (Equiv. People) As-Built Data
W ork Q ualityto D ate
ScheduledC om pletion
T im e
ExpectedC om pletion
T im eAvailab ility
o f P rerequ is ites
Perce ivedProgress
SchedulePressure
O ut-o f-SequenceW ork
M ora le
ExpectedH ours a t
C om pletion
H ours Expended
to D ate
Skill &Experience
H iring
Equiva lentS ta ff onPro ject
S ta ffingR equested
Progress
R ew orkD iscovery
TurnoverO rganizationa lS ize
C hanges
Staff
Productiv ity Q uality
AddedW ork O bsoleted
W ork
O vertim e
T im eR em ain ing
W ork ToBe D one
U ndiscoveredR ew ork
Know nR ew ork
W orkR eally D one
Assessing the Impact of Changes
Step 1: Recreate Program History (Project Plan + Changes)
13
+
-
TIME
Total Program Labor (Equiv. People)
Planned Program
Conditions
Company Actions
Customer-Responsible Events & Conditions
Simulation Model
As-Built Data Absent Customer Items
X
W ork Q ualityto D ate
ScheduledC om pletion
T im e
ExpectedC om pletion
T im eAvailab ility
o f P rerequ is ites
Perce ivedProgress
SchedulePressure
O ut-o f-SequenceW ork
M ora le
ExpectedH ours a t
C om pletion
H ours Expended
to D ate
Skill &Experience
H iring
Equiva lentS ta ff onPro ject
S ta ffingR equested
Progress
R ew orkD iscovery
TurnoverO rganizationa lS ize
C hanges
Staff
Productiv ity Q uality
AddedW ork O bsoleted
W ork
O vertim e
T im eR em ain ing
W ork ToBe D one
U ndiscoveredR ew ork
Know nR ew ork
W orkR eally D one
Changes removed in reverse chronological order to test cumulative impact.
Assessing the Impact of Changes
Step 2: Remove Direct Impact of Changes – “But for …”
14
+
- Assessing the Impact of Changes
Step 3: Difference is cost “but for” client impacts, including secondary impacts.
Budget
+ Other Sources of Cost Overrun
+ Direct Cost of Changes
+ Indirect Cost of Changes (1-10 X Direct)
---------------------------------------------
Total Project Cost
} Client Responsible Costs
15
+
-
% Impact on Cost Growth for the LHA Program
Excessive Schedule Pressure
23%
Out-of-Sequence Work 29%
Loss of Learning
3%
Inadequate Skill Level
3%
Reduced Drawing Quality
6% Delayed/Reduced
Quality of Prior Construction
14%
Other Labor Impacts
22%
D&D Results from Impact on Productivity & Fraction Correct of Unchanged Work
16
+
- The Navy’s starting position
“If you think you’re going to get 10 cents from us with this black box hocus-pocus simulation model, you’re nuts.”
But after a review by MIT System Dynamics Professors …
17
+
-
How it all got started: Ingalls Shipbuilding vs. US Navy
Case settled out of court for $447 million
Model-generated analysis was the basis for $200-300 million of the claim.
Source: Cooper, Kenneth G., 1980, Naval Ship Production: A Claim Settled and a Framework Built. Interfaces. 10(6)
(December), 20-36.
18
+
- Since the beginning ...
More than 50 contract disputes In excess of $4 billion in dispute, with
average recovery of 75% vs. 40% with traditional methods
All disputes have settled out of court, avoiding lengthy litigation
More than 150 “proactive” applications In excess of $25 million in consulting fees
Conservatively saved clients $5 billion on cost and schedule performance
19
+
-
Staff100
75
50
25
00 6 12 18 24 30 36 42 48 54 60
Time (Month)
Peo
ple
Staff : SD5 Changes Slip ScheduleStaff : SD5 ChangesStaff : SD5 Feasible Plan1
Managing Changes Proactively: Schedule Slip Mitigates Impact
Changes, Add Staff; Cost +28%
Changes, Slip Schedule; Cost +16%
20
Plan
7% of total scope directly redone (some multiplier inevitable b/c of technical interdependence)
+
-
Why, in face of changes, don’t firms consider schedule slip (or price for full impact)?
21
+
- Today’s Agenda
Managing Changes
• Case Example – Dispute Resolution • Case Example – Change Management
Project-to-Project Learning
• Case Example – Risk Management • Case Example – Bidding and Management
Broader issues – project priorities and market
interactions
22
+
- Case Example: Fluor Corporation
“Architect to Industry” ($20 billion annual revenue)
Highly centralized
Resistant to change
Business line specific cultures
Highly diversified Energy (production, refining, chemicals, power)
Commercial (hotels, office buildings, concert halls, food products)
Industrial (mining, pharma/bio, manufacturing, consumer goods)
Infrastructure (airports, hospitals, highways, high speed & light rail, ports)
Federal Government - DOD, DOE (nuclear fuel cycle), DOS
23
+
-
“Improving change impact management is vital to corporate performance”*
Waiting to recover for owner-induced change impacts via a dispute process is risky, expensive, and precludes impact mitigation
“Project changes represent the single largest source of project productivity impact” (2002 Fluor survey)
Changes do not increase project profits, but in fact are a source of profit reduction
*Greg Lee, Senior VP
24
+
- Change Impact Revenue Loss %
25
Proprietary Information
+
-
Learning Point: More and later changes create not just more impact, but disproportionately more impact.
Cost Impacts of Change – Results in Profitability Loss from Planned Rates
0%
10%
20%
30%
40%
50%
60%
70%
80%
0% 5% 10% 15% 20% 25%
Engineering Changes' Direct Impacts
More changes
Engineering Impact
(% of budget)
0%
10%
20%
30%
40%
50%
60%
70%
80%
0%
10%
20%
30%
40%
50%
60%
70%
80%
0% 5% 10% 15% 20% 25%
Engineering Changes' Direct Impacts
0% 5% 10% 15% 20% 25%
Engineering Changes' Direct Impacts
More changes More changes
Engineering Impact
(% of budget)
0%
10%
20%
30%
40%
50%
60%
70%
80%
0% 5% 10% 15% 20% 25%
Engineering Changes' Direct Impacts
Craft Impact
(% of budget)
0%
10%
20%
30%
40%
50%
60%
70%
80%
0%
10%
20%
30%
40%
50%
60%
70%
80%
0% 5% 10% 15% 20% 25%
Engineering Changes' Direct Impacts
0% 5% 10% 15% 20% 25%
Engineering Changes' Direct Impacts
Craft Impact
(% of budget)
Longer changes
2 mos.
4 mos.
6 mos.
8 mos.
10 mos.
26
Note: while these results are from a simulation model, an analysis of actual project results by Greg Lee of Fluor clearly demonstrated that projects which experienced more and/or longer change suffered reduced profitability from planned levels. “Changes” are not advantageous.
+
-
Fluor Now Uses SD Model as Basis for Change Management
Model set up and tailored to each engineering & construction project
Used to:
Foresee future cost & schedule impacts of project changes & events
Explain “secondary impact” to clients
Price changes to include the full and cumulative impact for appropriate cost recovery
Find mitigating actions to reduce client costs Avoid late changes
Resolve proposed changes quickly
Delay start/end of construction
27
+
- Forecast Accuracy -- Fluor
Quotes for project teams: “The tool simulated our staffing almost perfectly.”
Another team described how, contrary to expectations, the model foretold a different pace of engineering progress yet to come, an outcome that occurred just as simulated.
Yet another project team told of the “uncanny accuracy” from the simulation as the project progressed.
28
+
-
Cumulative project applications and benefits are growing strongly
Hundreds of project managers and planners have been trained in the ongoing internal use of the system.
29
+
-
Cost impact savings come from different mitigations
Over $1,600 million savings from…
(All schedule shifts tested were less than or equal to the amount of engineering delay caused by changes)
Shifted construction
start time
Reduced change
resolution time
Shifted construction end time
36%
33% 31%
30
+
- Award-Winning Work
Designation by Engineering Construction Risk Institute (ECRI) as “Industry Best Practice”
2009 System Dynamics Society Applications Award & 2011 Edelman Laureate
31
+
- Today’s Agenda
Managing Changes
• Case Example – Dispute Resolution • Case Example – Change Management
Project-to-Project Learning
• Case Example – Risk Management • Case Example – Bidding and Management
Broader issues – project priorities and market
interactions
32
+
- Project-to-Project Learning
Improve estimation and planning
Assist in risk identification and quantification
Determine effective processes and management actions
33
+
-
Useful Estimating & Planning Information
Scope (tasks done)
Rework
Undiscovered rework profile
Rework discovery profile
Pattern of productivity and fraction correct
34
+
-
35
Progress
Rework Discovery
Productivity Fraction Correct
Work To Be Done
Undiscovered Rework
Known Rework
Work Really Done
Data for estimation and control
But, the split cannot be determined as it is happening
Progress Can be Measured (with some delay?)
Rework discovery can be measured and used to estimate split after the fact
Cumulative effort
Cumulative work done
Staff and Overtime
+
-
Management Actions
Planned Program
Conditions (Plan)
Changes to Plan
Real Program
TIME
Total Program Labor (Equiv. People) As-Built Data
W ork Q ualityto D ate
ScheduledC om pletion
T im e
ExpectedC om pletion
T im eAvailab ility
o f P rerequ is ites
Perce ivedProgress
SchedulePressure
O ut-o f-SequenceW ork
M ora le
ExpectedH ours a t
C om pletion
H ours Expended
to D ate
Skill &Experience
H iring
Equiva lentS ta ff onPro ject
S ta ffingR equested
Progress
R ew orkD iscovery
TurnoverO rganizationa lS ize
C hanges
Staff
Productiv ity Q uality
AddedW ork O bsoleted
W ork
O vertim e
T im eR em ain ing
W ork ToBe D one
U ndiscoveredR ew ork
Know nR ew ork
W orkR eally D one
Risk Assessment
Analyze what happened on prior projects …
Large data bases of project performance histories; Scientific assessment of project performance: What happened and why, using “claims” process?
36
+
-
Original Work to Do Work Done
Undiscovered Rework
Progressx
Quality
Time to Discover Rework
Rework to Do Rework
Discovery
- +
Productivity
Effort Applied
Rew ork
Progress Rework
Generation +
+ +
Scope Growth
Changes
Data for Risk Assessment
Project Time
Project Time
Project Time
1
Delayed Availability of Staff
Project Time
1
Late Vendor Info; Extra Mtgs
Project Time
1
Uncertain Specs & Reqmnts.
Accumulate a data base of exogenous impacts
Fraction Correct and Complete
37
+
-
TIME
Total Program Labor (Equiv. People)
Planned Program
Conditions
Management Actions
Changes to Plan
Simulation Model
As-Built Data Absent Changes to Plan
X
W ork Q ualityto D ate
ScheduledC om pletion
T im e
ExpectedC om pletion
T im eAvailab ility
o f P rerequ is ites
Perce ivedProgress
SchedulePressure
O ut-o f-SequenceW ork
M ora le
ExpectedH ours a t
C om pletion
H ours Expended
to D ate
Skill &Experience
H iring
Equiva lentS ta ff onPro ject
S ta ffingR equested
Progress
R ew orkD iscovery
TurnoverO rganizationa lS ize
C hanges
Staff
Productiv ity Q uality
AddedW ork O bsoleted
W ork
O vertim e
T im eR em ain ing
W ork ToBe D one
U ndiscoveredR ew ork
Know nR ew ork
W orkR eally D one
Identifying Risks for Future Projects
Identify, quantify & “remove” direct impact of all changes
38
+
-
TIME
Total Program Labor (Equiv. People)
Planned Program
Conditions
Management Actions
Changes to Plan
Simulation Model
As-Built Data Absent Management Actions
X W ork Q uality
to D ate
ScheduledC om pletion
T im e
ExpectedC om pletion
T im eAvailab ility
o f P rerequ is ites
Perce ivedProgress
SchedulePressure
O ut-o f-SequenceW ork
M ora le
ExpectedH ours a t
C om pletion
H ours Expended
to D ate
Skill &Experience
H iring
Equiva lentS ta ff onPro ject
S ta ffingR equested
Progress
R ew orkD iscovery
TurnoverO rganizationa lS ize
C hanges
Staff
Productiv ity Q uality
AddedW ork O bsoleted
W ork
O vertim e
T im eR em ain ing
W ork ToBe D one
U ndiscoveredR ew ork
Know nR ew ork
W orkR eally D one
Identifying Effective Policies
?
Identify, Quantify & Remove
39
+
-
TIME
Total Program Labor (Equiv. People)
Planned Program
Conditions
Management Actions
Changes to Plan
Simulation Model
As-Built Data Absent Changes & Mgt Actions
X
W ork Q ualityto D ate
ScheduledC om pletion
T im e
ExpectedC om pletion
T im eAvailab ility
o f P rerequ is ites
Perce ivedProgress
SchedulePressure
O ut-o f-SequenceW ork
M ora le
ExpectedH ours a t
C om pletion
H ours Expended
to D ate
Skill &Experience
H iring
Equiva lentS ta ff onPro ject
S ta ffingR equested
Progress
R ew orkD iscovery
TurnoverO rganizationa lS ize
C hanges
Staff
Productiv ity Q uality
AddedW ork O bsoleted
W ork
O vertim e
T im eR em ain ing
W ork ToBe D one
U ndiscoveredR ew ork
Know nR ew ork
W orkR eally D one
Estimation: What Would Project Have Cost?
X Note for projects: this is the process you might use for a post-mortem SD model
40
+
- Example – Automotive Development
Post-mortem assessment of 11 projects using SD model – major sources of risk ranked in terms of impact on schedule and delivered:
1. Late information and/or changes
2. Resource availability Slow ramp up, lower peak, forced ramp down to meet budget
Inadequate skills mix
3. New processes, missing enablers, or new materials
4. Organization &/or geographic changes
5. Aggressive program assumptions Compressed timing, inadequate budget, lean allowance for
prototypes
41
Model used to price out mitigation savings for typical risks
+
- Today’s Agenda
Managing Changes
• Case Example – Dispute Resolution • Case Example – Change Management
Project-to-Project Learning
• Case Example – Risk Management • Case Example – Bidding and Management
Broader issues – project priorities and market
interactions
42
+
- The Peace Shield Program
Air defense system developed by Hughes AC
System dynamics model was used for:
Bid support and risk assessment;
On-going project management (assessing impact of process and staffing changes)
Program successfully completed on time and on budget
Post-project learning and policy assessment
43
+
-
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 0
200
400
600
800
1000
1200
1400
1600
1800
Peace Shield
Past Program
0
Cumulative Effort
Past Program
Peace Shield
Post-project learning ...
Staff Levels
44
+
-
What caused the differences between Peace Shield and Past Program?
l Differences in work scope?
l External Conditions?
l Management policies and processes?
And how can a company learn from these differences, and therefore
l Bid better?
l Plan better?
l Manage better?
45
+
-
1991 1992 1993 1994 1995 0
Data and Plans in mid-1993
Simulated
The same model with different exogenous “change” and “management” inputs accurately replicated Peace Shield
46
+
-
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 0
Simulated
Data
… and the Past Program
47
+
- Major external differences
Peace Shield --
Lower scope and fewer changes
Fewer vendor delays & hardware problems
Better hiring conditions (less delay)
(Some of these are sources of risk)
48
+
-
1 2 3 4 5 6 7 8 9 10 11 12 0
200
400
600
800
1000
1200
1400
1600
1800
Past Program
0
External Differences Removed
Cumulative Effort
Past Program
External Removed
Removing external sources of difference from the past program ...
49
+
-
1 2 3 4 5 6 7 8 9 10 11 12 0
200
400
600
800
1000
1200
1400
1600
1800
Past Program
0
Cumulative Effort
Past Program
External Removed
Removing external sources of difference from the past program ...
External Differences Removed
Peace Shield Schedules
50
+
- Management differences
Peace Shield --
Adopted teaming structure, including customer involvement in design reviews
Lower productivity, faster rework discovery
Different phase overlap & staffing strategy:
Assigned staff “rolling off” to QA
Delayed start of downstream work
Minimize “Errors on Errors” dynamic even if reported progress is lower
51
+
-
Rework Discovery Delay (Months)
Program Year
24 22 20 18 16 14 12 10
8 6 4 2 0
1 2 3 4 5 6 7 8 9 10 11
Past Program
Peace Shield
Shorter Rework Discovery Times on Peace Shield: Reduced “Errors on Errors” dynamic
• Teaming • Delayed Roll-off of Staff • Customer involvement in Design Reviews
52
+
-
OriginalWork to Do Work Done
UndiscoveredRework
Progressx
Quality
Time to Discover Rework
Rework toDo Rework
Discovery
- +
Productivity
EffortApplied
Rework
Progress ReworkGeneration
++ -
Lesson: Recognize the rework cycle and minimize its consequences
53
Fraction Correct and Complete
Increase
Reduce
Overestimates of Progress
Build Errors Into Downstream Work
+
-
1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 0
200
400
600
800
1000
1200
1400
1600
1800
Peace Shield
Past Program
0
Cumulative Effort
Past Program
Peace Shield
Removing management differences ...
External Differences Removed
Peace Shield Schedules
54
+
-
Better Hiring Conditions
3%
Fewer Vendor & Hardware Problems
19%
Delayed Ramp-Up 19%
Staff Assigned to QA 24%
Teaming & Other Improvements 13% Scope Differences &
Fewer Customer Changes 22%
External Conditions 44% Policies & Processes 56%
Where Did The Cost Improvement Come From?
These are “free” cost savings!
55
+
- Sources & Additional Reading
Ingalls: Cooper, Kenneth G., 1980, Naval Ship Production: A Claim Settled and a Framework Built. Interfaces. 10(6) (December), 20-36.
Fluor: http://www.kcooperassociates.com/files/SD_Paper_for_Reprint_V3.pdf
Peace Shield: James Lyneis, Kenneth Cooper, and Sharon Els, “Strategic Management of Complex Projects,”
System Dynamics Review, Fall 2001 (on the course site)
56
+
-
Why Do Organizations Seem So Poor at Learning Lessons From Prior Projects?
57
+
-
Project-to-Project Learning Requires a Framework (Model/Data/Process)
Analyze what happened on prior projects …
Rework cycle and feedback effects provide one framework for assessing dynamics similarities.
TIME
Total Program Labor (Equiv. People)
Planned Program
Conditions
Management Actions
Changes to Plan
Simulation Model
As-Built Data Absent Management Actions
W ork Q ualityto D ate
ScheduledC om pletion
T im e
ExpectedC om pletion
T im eAvailab ility
o f P rerequ is ites
Perce ivedProgress
SchedulePressure
O ut-o f-SequenceW ork
M ora le
ExpectedH ours a t
C om pletion
H ours Expended
to D ate
Skill &Experience
H iring
Equiva lentS ta ff onPro ject
S ta ffingR equested
Progress
R ew orkD iscovery
TurnoverO rganizationa lS ize
C hanges
Staff
Productiv ity Q uality
AddedW ork O bsoleted
W ork
O vertim e
T im eR em ain ing
W ork ToBe D one
U ndiscoveredR ew ork
Know nR ew ork
W orkR eally D one
?
58
+
-
59
You’re Uncomfortable With Quantifying All These Effects. What Are Your Options?
1. Ignore effects and estimate (simulate) impacts as if they did not exist But that’s the only value you know is wrong!
2. Use your experience/intuition/ “mental model” instead (no simulation) I.e., try to account for effects simultaneously in your head
that you can’t do individually in a computer model
3. Use computer model with educated estimates … Test sensitivity of results to exact values
Gather data (and calibrate where warranted)
Assemble a data base from prior projects
+
- Today’s Agenda
Managing Changes
• Case Example – Dispute Resolution • Case Example – Change Management
Project-to-Project Learning
• Case Example – Risk Management • Case Example – Bidding and Management
Broader issues – project portfolios, priorities and
market interactions
60
+
- Broader Issues
Issues in product portfolios
Market and Customer Dynamics -- setting the mission dimension as a part of corporate strategy
61
+
- Issues in Product Portfolios
Portfolio interactions --
staffing and other resources
technical interdependencies
What happens on one project has significant knock-on effects to other projects. Aggressive project assumptions (“inconsistent mission”) adversely affect more than the one project.
62
+
- Portfolio Resourcing Issues
Constraints on Shared Resources
Late and over-budget projects delay ramp-up of downstream projects
Shared resources (e.g. test facilities) can also create bottlenecks
Staff working simultaneously on multiple projects create inefficiencies and delays
Typically dealt with via exogenous inputs to single-project models, or via portfolio models
63
+
- Phasing of Project Staffing
Project Staffing
Time
Typical Plan
Project 1 Project 2
64
+
-
What happens when Project 1 fails to meet plan?
Project Staffing
Time
Typical Plan
65
+
- Phasing of Project Staffing
Project Staffing
Time
Typical Plan
Project 1
Project 2
66
+
- “Tipping Point” / “Fire-fighting” Research
Not only are resources constrained, but because of technical interdependencies, failure to adequately complete the first “project” causes more work and rework on the second “project”, etc.
In a situation of limited resources, this can lock the organization into a permanently low mode of performance
67
+
- Two Examples
Resources onConcept
Development
Completeness &Quality of Concept
Development
Resources Needed forDownstreamDevelopment
AvailableResources
+
+
-
-
68
Resources onDevelopment
Completeness & Quality ofDevelopment ("bugs" in
delivered product)
Resources Needed forClient Support and
Maintenance
AvailableResources
+
+
-
-
Temporary shortages of resources can …
… lock organization into “low” mode of “quality”
+
-
69
Temporary Resource Shortage
Can “tip” system to low quality mode.
+
- “Tipping Point” Case Examples
“Understanding fire fighting in new product development,” The Journal of Product Innovation
Management 18 (2001) [posted on the course site] Black, L. J. and N. P. Repenning. Why Firefighting Is
Never Enough: Preserving High-Quality Product Development. System Dynamics Review Vol. 17, No. 1, Spring 2001.
Rahmandad, H. and Weiss, D 2009. "Dynamics of concurrent software development." System Dynamics Review 25(3): 224-249.
70
+
- Corporate Strategy for the Project
Determining the fit of the project to business objectives (the “mission”)
features / scope
schedule milestones (time to market)
delivered quality (defects)
resources & budget
And the mix/timing of “projects” necessary to achieve corporate strategy
What gives when project gets in trouble?
71
+
- % Specifying 1st or 2nd Choice
72
What You Do?At 30% At 65%
Add People 40.8% 34.7%Longer Hours 24.3% 23.5%Intensity 21.4% 19.4%Slip 5.8% 11.2%Cut Scope 7.8% 11.2%Other 0.0% 0.0%Total 100.0% 100.0%
+
-
Control & Flexibility Actions Involve Tradeoffs
Project resources: “haste makes waste,” fatigue, experience dilution, …
Slip schedule: penalties, loss of market share (Lecture 13)
Cut “scope”: loss of market share, need to upgrade later (Lecture 13)
Ship with “bugs”/incomplete testing: loss of market share, diversion of resources to maintenance (Lecture 13)
73
+
-
Do you ever ship a project with known bugs in order to meet the schedule?
1. Yes
2. No
74
+
-
Why the focus on resource responses?
75
+
-
Consequences of Adding Resources to Meet Scope and Schedule Largely Constrained to Project Itself
WorkRemaining
ResourcesNeeded
Resources
+
+
-
Time Remaining
ScheduledCompletion Date
+
-
1. Add Resources
Product CostTrigger (changes, scope growth, etc.)
?
Responsibility of project manager?
76
+
-
Whereas slipping schedule impacts time-to-market
WorkRemaining
ResourcesNeeded
Resources
+
+
-
Time Remaining
ScheduledCompletion Date
+
-
1. Add Resources
Product Cost
FeasibleCompletion Date
+-
+
2. Slip Schedule
Time-to-Market+
?
Revenue and market share loss
77
+
-
Cutting scope impacts product capability and features
WorkRemaining
ResourcesNeeded
Resources
+
+
-
Time Remaining
ScheduledCompletion Date
+
-
1. Add Resources
Product Cost
FeasibleCompletion Date
+-
+
2. Slip Schedule
Time-to-Market+
Project Scope
+
+
+
Product Capabilityand Features
+
3. Cut Scope
?
78
Revenue and market share loss
+
-
Shipping when out of time means delivered quality low
WorkRemaining
ResourcesNeeded
Resources
+
+
-
Time Remaining
ScheduledCompletion Date
+
-
1. Add Resources
Product Cost
FeasibleCompletion Date
+-
+
2. Slip Schedule
Time-to-Market+
Project Scope
+
+
+
Product Capabilityand Features
+
3. Cut Scope
Bugs in ShippedProduct
+
Ship Product-
+
-
4. Ship WhenOut of Time
Because consequences of other control actions extend beyond the project and are difficult to measure, primary response is to add resources.
?
79
+
-
Evaluating market impacts requires a different model …
Revenue
Profit
Sales
Market Demand
Market SharePrice
++
++
Costs
+-
Revenue and Profit Drivers …
80
+
-
“Mission” Elements Affect Attractiveness
Revenue
Profit
Sales
Market Demand
Market SharePrice
++
++
ProductAttractiveness
ProductNewness
Quality
Scope
Costs
+
+
+
+
Time to Market
+
-
-
-
81
+
-
“Mission” elements have negative impacts as well …
Revenue
Profit
Sales
Market Demand
Market SharePrice
++
++
ProductAttractiveness
ProductNewness
Quality
ScopeDevelopement
Cost
CostsWarranty Costs
+
+
+
+
Time to Market
+
+
+
-
-
-
Mission Tradeoffs
?
82
+
-
Market Model with feedbacks through profit and budget
Revenue
Profit
Sales
Market Demand
Market SharePrice
++
++
ProductAttractiveness
ProductNewness
Quality
ScopeDevelopement
Cost
DevelopmentBudget
CostsWarranty Costs
+
+
+
+
Time to Market
+
+
+
-
-
-
+
Mission Tradeoffs; Product Portfolio
83
+
-
Selling System Dynamics (Modeling)
Must be a persistent and costly dynamic problem
Illustrate causes (use rework cycle and feedback examples)
Provide an example of use relevant to your organization
“SDM students are hereby granted permission to use any of my lecture slides for internal company presentations, with appropriate attribution.” James M. Lyneis, 11/1/2012
84
+
- SD Qualitative Insights -1
1. A feasible plan is essential, including:
Estimates of rework, undiscovered rework, and delays in discovering that rework
Estimates of productivity loss dealing with rework
Adequate buffers and reserves for rework
[Rework increases with project uncertainty and complexity]
85
+
- SD Qualitative Insights – 2
2. A feasible plan recognizes the “iron triangle”; there will be multiple “feasible” plans depending on priorities.
3. Tradeoffs in the plan can often be improved by changes in project structure and organization to reduce rework and delays in discovering rework.
86
+
- SD Qualitative Insights – 3
4. Attempts to achieve an infeasible plan via project control actions lead to “vicious circle” side effects which increase project cost and duration.
On complex projects, these costs usually exceed the “direct” costs of infeasibility
5. Project “changes,” and risks which materialize, are fundamentally the same as an infeasible plan. (Lecture 13)
87
+
- SD Qualitative Insights – 4
6. Project managers need buffers and/or flexibility (e.g., slip schedule, cut scope, ship with “bugs”) to respond to changes and uncertainties. These have costs that need to be evaluated; the importance of different tradeoffs differs by project. (Lecture 13)
7. The costs of project control can be minimized by understanding the sources of the vicious circles. The timing, magnitude, and duration of different controls affects performance.
88
MIT OpenCourseWarehttp://ocw.mit.edu
ESD.36 System Project ManagementFall 2012
For information about citing these materials or our Terms of Use, visit: http://ocw.mit.edu/terms.