enterprise modeling
Post on 23-Feb-2016
64 Views
Preview:
DESCRIPTION
TRANSCRIPT
Enterprise Modeling
Saad El BeleidyPeyman Jamshidi
Jared KovacsGabriel Lewis
Improving the Analysis of Alternatives (AoA) Process for IT Solutions Government
Contract Proposals
Presented at:
Contents
Context and Stakeholder AnalysisProblem and NeedDesign AlternativesSimulation DesignResults and Conclusions
2
Contents
3
Context and Stakeholder AnalysisProblem and NeedDesign AlternativesSimulation DesignResults and Conclusions
IT Solutions Government Contractors
4
Project Sponsor: Civilian and National Security (CNS) Division, Vangent, Inc.
Proposal Development is critical to gaining revenueAoA: develop complex IT solution, crucial to proposal15-25 Proposals per year, of varied sizesFederal Contract Spending decreased $40 billion from 2009 to 2011
1 2 3 4 5 6 7 8 90
0.4
0.8
AoA Duration (Weeks)
Like
lihoo
d
2009 2010 2011$500
$530
$560
Fiscal Year
Fede
ral C
on-
trac
t Sp
endi
ng
(Bill
ions
)
Adapted from the Federal Times, 2011
5
Acquisitions CommitteeSolicitation Proposals
Selected Solution
Government Contractor
Bid Decisio
n
Government Entity
Requirements
Technical Solution Development
Proposal Writing
Budget & Management
Technical Solution Development / AoA
Process Level 1Proposal Development
Process
Highly competitive: 5-10 competitors for each proposal
Process Level 2 Top Level AoA Process
6
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Solicitation
Ranked List of
Alternatives
High-level “Decision and Analysis Resolution” (DAR) model provided by Vangent sponsors
0%20%40%60% 20% 10% 30% 40%
Mea
n Du
ra-tio
n (%
AoA
)
Task Category Classification1. Labor Intensive:
Little expertise required: parsing, documenting.Time Variability = Low (±22%)
2. Decision Making:Expert required: difficult decisions, judgment calls.Time Variability = High (±54%)
3. Experience Recall:Expert required: subjectively making judgments based on previous experience. Time Variability = Medium (± 44%)
4. Networking:Personal dialogue and collaboration with co-workers and others.Time Variability = High (±100%) 7
Process Level 2Top Level AoA Process
8
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Solicitation
Ranked List of Alternatives
High-level “Decision and Analysis Resolution” (DAR) modelprovided by Vangent sponsors
Process Level 3Phase 1: Define Problem Domain
9
Data acquired via knowledge elicitation from stakeholder SMEs
Task Category
% of Phase
Labor Intensive 25%
Decision Making 23%
Experience Recall 21%
Networking 31%
Process Level 2Top Level AoA Process
10
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Solicitation
Ranked List of Alternatives
High-level “Decision and Analysis Resolution” (DAR) modelprovided by Vangent sponsors
Process Level 3Phase 2: Define Evaluation Criteria
11
Data acquired via knowledge elicitation from stakeholder
SMEs
Task Category
% of Phase
Labor Intensive 42%
Decision Making 20%
Experience Recall 18%
Networking 20%
Process Level 2Top Level AoA Process
12
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Solicitation
Ranked List of Alternatives
High-level “Decision and Analysis Resolution” (DAR) modelprovided by Vangent sponsors
Process Level 3Phase 3: Explore Alternate
Solutions
13
Data acquired via knowledge elicitation from stakeholder
SMEs
Task Category
% of Phase
Labor Intensive 36%
Decision Making 15%
Experience Recall 13%
Networking 36%
Process Level 2Top Level AoA Process
14
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Solicitation
Ranked List of Alternatives
High-level “Decision and Analysis Resolution” (DAR) modelprovided by Vangent sponsors
Process Level 3Phase 4: Evaluate Solutions
15
Task Category
% of Phase
Labor Intensive 38%
Decision Making 21%
Experience Recall 23%
Networking 18%
Data acquired via knowledge elicitation from stakeholder
SMEs
Current Issues in AoA1. Limited and variable availability
of past research due to proprietary restrictions
2. Limited and variable applicability of past research once gained
3. Variable difficulty of AoA—need for past research
16
0% 50% 100%0
0.25
0.5
Relative Expected Availabil-ity
Prob
abili
ty
0% 50% 100%0
0.25
0.5
Relative Expected Appli-cability
Prob
abili
ty
Current Issues in AoA4. Quality of AoA suffers from
the information’s lack of availability and applicability
5. One employee handles entire AoA, so potentially parallel processes are conducted in series
6. Time spent for AoA results is an entirely overhead cost
17
0% 50% 100%0%
100%
Relative Expected Availabil-ity
AoA
Out
-pu
t Re
la-
tive
Qua
l-ity
0% 50% 100%0%
50%100%
Relative Expected Appli-cability
AoA
Out-
put
Rela-
tive
Qua
l-it
y
1 2 3 4 5 6 7 8 90
0.4
0.8
$32,000
$16,000
$6000
AoA Duration (Weeks)Pr
obab
il-it
y
Key Stakeholder GoalsSolutions ArchitectsPerform Analysis of Alternatives– Maximize productivity– Avoid overtime– Maximize labor rewards
Capture ManagersManage the transition from opportunity discovery to contract award.Oversees bid strategies, pricing, and teaming– Maximize probability of winning contracts
• Increase AoA discriminability and quality
Proposal ManagersDevelop and manage the proposal plan and schedule– Maximize proposal/AoA throughput 18
Key Stakeholder Interactions
19
InteractionsProposal DemandSolution Demand
Tension:Managers and
Solutions ArchitectsLimited time and personnel resources to conduct AoAs.
Contents
20
Context and Stakeholder AnalysisProblem and NeedDesign AlternativesSimulation DesignResults and Conclusions
Problem StatementDuring a time of national economic downturn, federal contract spending cuts have led to a decrease in available contract revenue and an increase in competition between government contractors. These factors have increased the time sensitivity of proposal development, specifically in the AoA process.
21
Need Statement
There is a need for Analysis of Alternatives process improvements to reduce the mean time duration by at least 33%, and the variability by 25%, while maintaining or increasing AoA proposed solution quality and keeping maximum costs below $100,000 per AoA.
22
Need validated with key stakeholders
Contents
23
Context and Stakeholder AnalysisProblem and NeedDesign AlternativesSimulation DesignResults and Conclusions
Design Alternative Approach
Optimize AoA Staffing Levels
– Target Parallel Tasks– Reduce mean time duration
Information Management System
– Target efficiency increases– Reduce duration and variability
24
AoA Phase Parallel Tasks
Define Problem Domain
6Define Evaluation Criteria
4Explore Alternate Solutions
5Evaluate Solutions 4
Task Category % AoA Variabilit
yLabor Intensive 35.2% Low
±22%Decision Making 19.2% High ±
54%Experience Recall 26.3% Med ±
44%Networking 19.2% High
±100%
Optimize AoA Staffing Levels
1 Additional Solutions Architect collaborates to conduct AoA
Reduce mean time duration of AoA– Additional resource to conduct parallel tasks– Increases size of social network– Potential for conflict in making decisions
Cost: Approx. $200,000 salary including benefits per yearFive year cost: $1,000,000
25
Task Category
Labor Intensive
Decision Making
Experience Recall
Networking
Efficiency Gain --- -10% --- +10%
Information Management Alternatives
26
Implementing a File Management System– Database searching capability improves availability of information in the AoA– Increases efficiencies of task categories, improves re-use of past research
Implementing a Content Management System– Enhanced capabilities greatly improves availability of information in the AoA– Increases efficiencies of task categories, improves re-use of past research
Maintaining a Sanitized Repository– Adds new content to the information pool– Changes high-variability task categories to low-variability categories
File Management System
Benefits– Integrates with current system– Organized file structure– Promotes collaboration– Easily scalable
Drawbacks– Requires permissions for file
access– Limited search functionality
Cost: IntravationInitial Cost: 100 User LicenseGDIT(Parent Company) has 25 user license; $1000 one time fee per active user
First Year Cost: $77,000Annual Maintenance: $1600Five Year Cost: $83,000
27
Task Category
Labor Intensive
Decision Making
Experience Recall
Networking
Efficiency Gain +10% +10% +15% +5%
Content Management System
Benefits– Robust searching and indexing– Preconfigured user roles based upon
content–access needs– Authentication, check in/out,
tracking– Workflow management
Drawbacks– High complexity
• High learning curve• Expensive Technical Support
Cost: DocumentumInitial Cost:100 User License Cost: $110,665
System Cost:First Year Cost: $129,000Annual Maintenance: $19,000Five Year Cost: $203,000
28
Task Category
Labor Intensive
Decision Making
Experience Recall
Networking
Efficiency Gain +15% +15% +20% +10%
Sanitized Document Repository
Benefits– Quality of information
improvement• Promotes availability• Increases applicable content
– Virtually eliminates security risks
– Provides quicker access to data– Minimal technical support
Drawbacks– Low initial benefit
CostEstimate 12 labor hours per AoA
Five Year Cost: $150,000
29
Task Category
Labor Intensive
Decision Making
Experience Recall
Networking
Efficiency Gain +15% +10% +10% +5%Variability Low High Med High
% of AoAOriginal New
35% 40% 20% 17% 19% 17% 26%
Contents
30
Context and Stakeholder AnalysisProblem and NeedDesign AlternativesSimulation DesignResults and Conclusions
Simulation DesignMonte Carlo Discrete Event Simulation of the AoA1000 ReplicationsEach Replication covers 25 Proposals (1 year)
Key Model Assumptions:1. Solutions Architects work on one task at a time2. Only one proposal is being worked on at any given time3. The four task categories adequately capture AoA time
consumption4. All tasks are of equal importance to the quality of AoA
output31
Simulation Design
32
Simulation Model
AoA Process Definition
Quality Metric
AoA Duration
Variability
AoA Mean Duration
Number of Technologie
s
AoA Difficulty
Information Availability
Information Applicability
Task Category
Efficiencies
Simulation Video
: Task Categories (Labor Intensive, Decision Making, Experience Recall, Networking)
: Inherent Task Delay : Task Category Efficiency Index (% Expected Time): Task Category Weight (% of task with specific task category): Task Category Variability Factor (RV): Technology Variability (RV): Number of Technologies/AoAs in the Proposal
Task Time Delay Equation
34
𝐷𝑒𝑙𝑎𝑦=∑𝑇𝐶
[𝐷∗𝐸∗𝑊∗𝑉 ∗𝑇 𝑉
𝑁 ]
Task Quality Metric Equation
35
: Availability Factor (RV): Applicability Factor
(RV): Stakeholder weights
Design of Experiment Matrix
36
A1: Staffing LevelsA2: Sanitized RepositoryA3: File ManagementA4: Content Management
Run Configuration AlternativesA1 A2 A3 A4
1 Baseline - - - -2 A1 x - - -3 A1, A2 x x - -4 A1, A2, A3 x x x -5 A1, A2, A4 x x - x6 A1, A3 x - x -7 A1, A4 x - - x8 A2 - x - -9 A2, A3 - x x -10 A2, A4 - x - x11 A3 - - x -12 A4 - - - x
Contents
37
Context and Stakeholder AnalysisProblem and NeedDesign AlternativesSimulation DesignResults and Conclusions
A1 A1, A2 A1, A2, A3
A1, A2, A4
A1, A3 A1, A4 A2 A2, A3 A2, A4 A3 A40.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
36.53%43.91%
49.47% 51.66%
43.08% 45.42%
11.15%
20.10%23.38%
10.11%13.97%
Alternative Configurations
Perc
ent D
ecre
ase
in A
oA
Mea
n Ti
me
Dur
atio
n
38
33.00%
A1: Optimized Staffing Levels; A2: Sanitized Repository;A3: File Management; A4: Content Management
Simulation ResultsPercent Decrease in Mean Duration
A1 A1, A2 A1, A2, A3
A1, A2, A4
A1, A3 A1, A4 A2 A2, A3 A2, A4 A3 A40.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
34.38%37.50%
50.00% 50.00%
37.50% 37.50%
6.25%
12.50%
18.75%
6.25% 6.25%
Alternative Configurations
Perc
ent D
ecre
ase
in A
oA
Tim
e Du
ratio
n Va
riab
ility
Simulation Results
Percent Decrease in Duration Variability
39
25.00%
A1: Optimized Staffing Levels; A2: Sanitized Repository;A3: File Management; A4: Content Management
Simulation ResultsQuality Metric
40
Baseline
A1 A1, A2
A1, A2, A3
A1, A2, A4
A1, A3
A1, A4
A2 A2, A3
A2, A4
A3 A40.7
0.72
0.74
0.76
0.78
0.8
0.82
0.84
Alternative Configurations
Qua
lity
Met
ric
A1: Optimized Staffing Levels; A2: Sanitized Repository;
A3: File Management; A4: Content Management
Using a Sanitized Repository is the only proposed alternative that directly affects quality.
~10% Increase in Quality
0.74Baselin
e
Utility
Mean Duration
Duration Variability
AoA Output Quality
Solution Quality
41
Stakeholders’ Utility Function
Value Hierarchy obtained via Stakeholder values elicitation weighted with the Swing Weight Method
.238 .048 .143 .571
Alternative Utility Ranking
42
RankAlternative Configuratio
nUtility
1 A1, A2, A4 4.252 A1, A2 3.953 A1, A2, A3 3.904 A2, A4 3.905 A2 3.546 A1, A4 3.497 A2, A3 3.288 A4 3.189 A1, A3 3.1410 A1 3.0911 Baseline 2.5012 A3 2.37
A1: Optimized Staffing Levels A2: Sanitized Repository A3: File ManagementA4: Content Management
Meet The Need
Don’t Meet The Need
43
Sensitivity AnalysisPercent Change in Criteria Weight Necessary to Change Utility Rank
Alternative Configuration
s
Utility Function Criteria
Mean Time Usability Integrability
A1, A2 +33% +41% +69%A1, A2, A3 +34% +48% +119%
A2, A4 +60% +34% +60%
-$5,000.00 $15,000.00 $35,000.00 $55,000.00 $75,000.00 $95,000.001
1.5
2
2.5
3
3.5
4
4.5
5
Baseline
A1
A1, A2 A1, A2, A3A1, A2, A4
A1, A3A1, A4
A2A2, A3
A2, A4
A3
A4
Cost per AoA
Util
ity
44
Cost-Benefit Analysis
A1: Optimized Staffing Levels; A2: Sanitized Repository;A3: File Management; A4: Content Management
Meet the Duration and Variability Reduction NeedDo Not Meet the Duration and Variability Reduction Need
-$5,000.00 $15,000.00 $35,000.00 $55,000.00 $75,000.00 $95,000.001
1.5
2
2.5
3
3.5
4
4.5
5
Baseline
A1
A1, A2 A1, A2, A3A1, A2, A4
A1, A3A1, A4
A2A2, A3
A2, A4
A3
A4
Cost per AoA
Util
ity
45
Cost-Benefit Analysis
A1: Optimized Staffing Levels; A2: Sanitized Repository;A3: File Management; A4: Content Management
Meet the Duration and Variability Reduction NeedDo Not Meet the Duration and Variability Reduction Need
More Desirable
Closest to Desirable Region
46
RecommendationsOptimize Staffing Levels and Maintain a Sanitized Repository (A1, A2)
– Percent Reduction in Time Duration: 44%– Percent Reduction in Duration Variability: 38%– Percent Increase in Quality: 10%– Total Utility: 3.95– Max Expected Cost per AoA: $50,000– Total Implementation Cost: $230,000/year
47
RecommendationsOptimize Staffing Levels, Maintain a Sanitized Repository, and Implement a Content Management System (A1, A2, A4)– Percent Reduction in Time Duration: 52%– Percent Reduction in Duration Variability: 50.00%– Percent Increase in Quality: 10%– Total Utility: 4.25– Maximum Expected Cost per AoA: $78,000– Total Implementation Cost: $111,000 plus
$249,000/yr
(Potential Value, if Cost Reduced)
Questions?
48
Backup Slides
49
Project Plan
50
WBS – Top LevelEnterprise Modeling
Project
1.0 Project Definition
2.0 Requirements Development
3.0 Solution Development
4.0 Modeling and Testing
5.0 Results Analysis
6.0 Communications and Management
Syst 495-Spring 2012
Context
Syst 490-Fall 2011 51
Output of AoA: Ranked Set of Alternatives
AoA output for “Case Management Platforms” (Vangent, Inc)
Vendors Considered: Custom Developed Siebel IBM Microsoft
Total Scores
Syst 495-Spring 2012 52
Stakeholders’ Value Hierarchy
Value Hierarchy obtained via Stakeholder values elicitation weighted with the Swing Weight Method
AoA Process Improvement
Objective
Reduce Mean Time
Maximize Usability
Reduce Time Variability
Maximize Integrability
Maximize Tailorability
Maximize AoA Quality
Maximize Tech Support
Maximize Scalability
.238 .048 .143 .024
.071.119.167.190
Information Management Alternatives
Syst 495-Spring 2012 53
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Retrieve Past Proposal Requirements
Maintain List of SME’sRetrieve Past Alternative Material
Retrieve Past Criteria from similar proposals
Provide reference of past analyses
Optimizing Staffing Levels Alternative
Syst 495-Spring 2012 54
Time ReductionParallel Processes
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Percent Decrease
55Syst 495-Spring 2012
Baseline
A1 A1, A2
A1, A2, A3
A1, A2, A4
A1, A3
A1, A4
A2 A2, A3
A2, A4
A3 A40.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
Mean Time (Filled) Variability (Hollow)A1: Optimized Staffing Levels; A2: Sanitized Repository;
A3: File Management; A4: Content Management
33%
25%
Alternatives Efficiency Summary
Syst 495-Spring 2012 56
Alternative
Labor Intensiv
eDecision Making
Experience Recall
Networking
Baseline --- --- --- ---Optimizing Staffing Levels
--- -10% --- +10%Sanitized Repository
+15% +10% +10% +5%File Management
+10% +10% +15% +5%Content Management
+15% +15% +20% +10%
Alternatives’ Efficiency Indexes
Syst 495-Spring 2012 57
Base-line
A1 A1, A2 A1, A2, A3
A1, A2, A4
A1, A3 A1, A4 A2 A2, A3 A2, A4 A3 A40.6
0.7
0.8
0.9
1
1.1
1.2Labor Intensive Decision Making Experience Recall Networking
Alternative Configurations
Effici
ency
Inde
x
A1: Optimized Staffing Levels; A2: Sanitized Repository;A3: File Management; A4: Content Management
Syst 495-Spring 2012 58
Acquisitions Committee
Solicitation Proposal
Selected Solution
Government Contractor
Bid Decisio
n
Government Entity
Requirements
Technical Solution Development
Proposal Writing
Budget & Management
Technical Solution Development
Process Level 1Proposal Development
Process
Proposal Development is critical to gaining revenueHighly competitive: 5-10 competitors for each proposal
Syst 495-Spring 2012 59
Process Level 2Technical Solution
Development
59
Alternatives
AnalysisIntegration of Assets
SolicitationRanked List
of Alternatives
Assets for Proposal
Solution to be Proposed
Analysis of Alternatives
(AoA)
Model developed based on team research and knowledgeelicitation from Vangent sponsors
Process Level 3 Top Level AoA Process
Syst 495-Spring 2012 60
Define the Problem Domain
Define Evaluation
Criteria
Explore Alternate Solutions
Evaluate Solutions
Solicitation
Ranked List of Alternatives
High-level “Decision and Analysis Resolution” (DAR) modelprovided by Vangent sponsors
top related