1. DEVELOPMENT OF UNCERTAINTY FACTORS FROM
HISTORICAL NASA PROJECTS
Dr. Tahani Amer, OE/IPAO NASA-HQ
Dr. Will Jarvis, OE/IPAO NASA-HQ
Dr. Ariel Pinto, Old Dominion University
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage.
2. Project Background
• Uncertainty analysis should be performed to capture the
program risks.
• NASA has been using available uncertainty factors from
Aerospace, Air Force, and Booz Allen Hamilton.
• NASA has no insight into the development of these factors,
which can lead to unrealistic risks in most NASA projects.
• SRB analysis requires cost growth factors that are
transparent and based on recent NASA experience.
• Better estimation and prediction of missions cost is
needed early on.
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 2
3. Data Collection Methodology
• NASA Fiscal Year Budget Estimates
• Cost Analysis Data Requirement (CADRe)
• Government Accountability Office Reports
• NASA Personal- IPAO/CAD/SMD
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 3
4. Data Collection Model
Budget Report
CADRe Report
Analysis- Recheck
Data Set
Assess Assess Assess
Analysis- Recheck
GAO Report
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 4
5. Source of Data
FY 2009 Budget
February 4, 2008
Amer DENG 5
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 5
6. Data Sampling Criteria
• Difficulty in collecting accurate data – limited dataset
• Criteria
– Science Mission Projects
– Developmental Phase
– Only cost data
– Use CARDe data as final for completed Missions
– For active missions: used in search of credible earliest available
and latest available estimates. Errors probably exist because
some projects are still in development and continue to evolve
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 6
7. Summary of NASA Missions Investigated
Active
Completed Missions Missions
NEAR MER HESSI EOS-Aqua AIM JUNO
LUNAR MRO GALEX EOS-Aura DAWN AQUARIUS
PROSPECTOR FAST SWIFT LANDSAT-7 PHOENIX LDCM
GENESIS SWAS GRACE TRMM GLAST NPP
MESSENGER TRACE CLOUDSAT TIMED KEPLER GPM
MARS WIRE CALIPSO GRAVITY SDO MMS
PATHFINDER DS-1 PROBE B WISE JWST
ACE
STARDUST EO-1 THEMIS NEW MSL
FUSE
CONTOUR SIRTF HETE-II HORIZONS GLORY
IMAGE STEREO GRAIL
DEEP IMPACT SORCE LRO
MAP
MGS ICESAT OCO
MCO/MPL
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 7
8. Initial & Final Cost for 50 Completed
Projects and Cost Growth Percentage
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 8
9. 60 NASA Missions - Historical Data
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 9
10. Mission Cost Growth Histogram
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 10
11. NASA & DOD Distribution
14
12
10
Number of Project
8
6
4
2
0
-21-0% 0 -10% 10-20% 20-30% 30-40% 40-50% 50-60% 60-70% >70%
Permission from McCrillis, DOD
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 11
12. 60 NASA Missions
14
12
10
Number of Project
8
6
4
2
0
- 21-0% 0 -10% 10-20% 20-30% 30-40% 40-50% 50-60% 60-70% >70%
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 12
13. Reasons for Cost Growth
• The major categories for cost growth:
– Overly optimistic and unrealistic initial cost estimates
– Project instability and funding issues
– Problems with development of instruments and other spacecraft
technology
– Launch service issues
– “Stuff” Happens
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 13
14. Uncertainly Factors from non-NASA Missions
Aerospace Air Force
Level Optimistic Most Likely Pessimistic
Low 0.95 1 1.1
Low Plus 0.96 1 1.23
Moderate 0.97 1 1.36
Moderate Plus 0.98 1 1.49
High 0.98 1 1.61
High Plus 0.99 1 1.74
Very High 1 1 1.87
Very High Plus 1 1 2
Booz Allen Hamilton
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 14
15. NASA Data & Air Force UF
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 15
16. NASA Data & Aerospace UF
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 16
17. NASA Data & BAH UF
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 17
18. Development of Distribution-54
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 18
19. Development of Distribution-44
0.12
0.11
0.1
0.09
0.08
0.07
PDF
0.06
0.05
0.04
0.03
0.02
0.01
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
Cost Growth (%)
Lognormal
µ= 3.03
F(X) = σ= 1.08
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 19
20. More Test Runs
• All data including Project A & B (A- Data), 54 missions.
• All data excluding Project A & B (B- Data), 52 missions.
• All completed missions (C-Data), 44 missions.
• All data less than 100% cost growth (D- Data).
0.12 Type of Data µ =Mean σ= Std
0.11 Dev.
0.1
0.09
A - Data 3.01 1.18
0.08
20.3 %
0.07
B - Data 2.92= 1.11
PDF
0.06
18.5 %
0.05
C - Data 3.03 1.08
0.04
0.03
Mean=37.08
0.02
Median= 20.7
0.01 Std. Dev.= 55.14
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150
D-Data 2.85 1.07
Cost Growth (%)
17.3%
October 7, 2011 Lognormal
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 20
21. Selected Distribution
• The lognormal distribution is the most suitable distribution that
the cost estimator can use to perform an uncertainty analysis for
the NASA data.
• The quality of the fit of two tests:
– The Kolmogorov-test 0.032
– The Anderson-Darling test 0.028
0.024
0.02
PDF
0.016
0.012
0.008
0.004
0
0 50 100
Cost Growth (%)
Lognormal (1.08; 3.03)
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 21
22. Recommended UFs
Uncertainty
Level of Risk Factors
No- Risk
Adjusted Conservative Aggressive
Moderate (10-30%) 1 1.1 1.3
High (30-75%) 1 1.3 1.75
Very High (>75%) 1 1.75 2.5
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 22
23. UFs w/Conservative High Risk Level
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 23
24. UFs w/Aggressive Moderate Risk Level
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 24
25. Guideline to UFs
• Cost Analyst Judgment
• Mission Complexity: Number of Instruments,
Technical Development
• Heritage Level
• Partners Performance
• Mission Visibility: International & Commercial
• Similarity of Missions
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 25
26. Comparison of UFs
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 26
27. UFs ------Test Case
– Project A---
• Development Cost as of 2006 ~ $969 M
• Development Cost as of 2011 ~ $1.8 B
• Apply NUFs:
– Make the following assumptions:
» High Risk Mission (10 instruments, new Technology, Planetary)
» Aggressive Risk
» Select a range from NUFs, Table 20, 1.75-2.5
• Expected development estimate range: 1.69- $2.42 B
• Other factors estimates:
– AF (1.61- 1.74) $1.56B- $1.69 B, which is lower than the current
actual development cost.
– Aerospace (2.05-2.5) $1.99- $2.52 B, which is much higher then
expected estimate from the low range. Note: LCC is @$2.5B.
– BAH (1.9-2.10) $1.84- $2.03 B, which is might be closer to the
expected development cost . Note: see section 3.3 for these
factors interpolation from BAH table.
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 27
28. UFs Test Case
970 1.2 1.4 1.8 2.2 2.6 2.8
Current EST.
$1.69- $2.42B
. UF
$1.56B- $1.69B AF
$1.99- $2.52B AS
$1.84- $2.03B BAH
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 28
29. Conclusions
• This research provides an important contribution to the
discipline of cost estimating.
• It has developed a solid database of actual cost growth
history and adds some statistical rigor to the derivation
of cost growth factors based on the data.
• Thoughtful execution and effective monitoring of the
NUFs will improve the cost estimating.
• Additionally, it is expected that this work will be
referenced often for independent cost estimates,
correction of advocacy-bias in project-generated
estimates, and other programmatic work.
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 29
30. Point Taken
• The greatest benefit of this research is providing a
proactive approach to make a quick cost estimating as
in back of envelope by:
– Program Managers
– Decision Makers
– Congressional Staffer
– Cost analysts
– Awareness and reference to Standard Review Board
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 30
31. Future Work
• Increase the number of historical missions captured in
the database.
• Validate UFs with more available actual cost data.
• Test the UFs for missions in various life-cycle phases.
• Develop a more adaptable tool for NUFs.
• Accept the UFs by the cost analysis’ community for
implementation.
• Develop Agency standard.
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 31
32. Backup
NASA Program Management Challenge 2012 This briefing is for status only and does not represent complete release of usage. 32