Military Aircraft Cost Database

1
Military Aircraft Cost Database Archive and
Retrieval
Michael J. Clark Air Force Cost Analysis
Agency Saroja Raman NAVAIR Cost Department
June 2002
2
OUTLINE

• Background
• Structure
• Demo
• Validation/Verification
• Material Cost Analysis
• Objective/Goal
• Process
• Results
• Future MACDAR Endeavors

3
Description

• Military Aircraft Cost Database Archive and
  Retrieval (MACDAR)
• Cost database which currently includes five
  military fixed wing aircraft
• F-14
• F-15
• F-16
• F/A-18
• A/V-8B
• Time span from early 1970s through early 1990s
• Tried to capture A/B variant through E/F variant

4
Purpose

• Provide a central location for military aircraft
  cost data down to a WBS level 4 that is
  consistent across all platforms.
• Use data for regression analysis and CER
  development in support of JSF, F-22, and UCAV
  production cost estimates.

5
History

• Air Force Cost Analysis Agency (AFCAA) initiated
  MACDAR project in 1991.
• Team of government and contractor personnel
  collected low-level cost data.
• Labor and material cost data collected for EMD as
  well as Production.
• Organized data into excel workbooks.
• Included evaluation of Foreign Military Sales
  (FMS).
• Validation Verification (VV) effort performed
  by NAVAIR and AFCAA began in 1998.
• Labor databases (recurring and non-recurring)
  have been validated and are currently in use.
• VV of recurring material databases to be
  completed by end of FY02.

6
Structure

• Five platforms included in MACDAR
• F-14, F-15, F-16, F/A-18, AV-8B
• Four excel workbooks for each platform
• Example
• F-14 rec. labor hrs
• F-14 non-rec. labor hrs
• F-14 rec. matl
• F-14 non.rec. matl
• Four workbooks do NOT capture total cost of
  Aircraft.

7
Structure

• Standard MACDAR WBS Structure developed, to
  provide consistency across platforms.
• Total Aircraft System
• Air Vehicle
• Airframe
• fuselage, wing, empennage, etc.
• Subsystems
• electrical, environmental control, hydraulics,
  etc.
• Propulsion
• Avionics H/W
• radar, navigation, controls displays, etc.
• Avionics S/W
• Mapping Schemes developed for each contractor.
• Can be changed by user

8
Structure

• Standard MACDAR Functional Categories developed,
  to provide consistency across platforms.
• Mapping Schemes developed for each contractor.
• Can be changed by user.

Material Databases Engineering
Tooling General Material
Subcontractor
Labor Databases Engineering Tooling
Manufacturing QA
9
Structure

• Each workbook contains labor hour data or
  material cost data organized
• by year
• by WBS
• by functional category (engr, tooling, mfg, gen.
  mat etc.)
• Raw data, mapping schemes and normalization sheet
  included in every workbook.
• Workbooks also include some technical/programmatic
  information.
• For example weight and quantity
• Pivot tables provide output. Pivot on a WBS
  element and a functional category and receive
  cost data for EMD and Production years.

10
DEMO
11
MACDAR DEMOLABOR
12
MACDAR DEMOLABOR
13
MACDAR DEMOLABOR
14
MACDAR DEMOMATERIAL
15
MACDAR DEMOMATERIAL
16
MACDAR DEMOMATERIAL
17
Validation Verification

• Ensure cost data accurate and database
  functioning properly
• Validate MACDAR data against Contractor Cost Data
  Reports (CCDRs)
• Identify missing costs
• Verify databases operating properly
• Ensure pivot tables provide correct output
• Perform corrections
• Correct database operating errors
• Use various sources of data to fill in missing
  costs

18
Challenges

• Purchased equipment and avionics data difficult
  to obtain, at low levels.
• Difficult to validate data from the 1970s.
• Five databases to validate, each with its own
  unique problems.
• Past CCDRs had their own inherent problems.

19
Accomplishments

• All labor databases, both recurring and
  non-recurring, have been validated.
• Three out of five, recurring material databases,
  have been validated and verified.
• All five recurring material databases to be
  completed by the end of FY02.

20
MATERIAL COST DATA LEARNING CURVE ANALYSIS
21
Objective

• Obtain a better understanding of military
  aircraft material costs in order to more
  accurately predict material costs for future
  military aircraft.

22

Goal

• Obtain learning curve slopes for various aircraft
  material costs.
• Start at lowest level possible (WBS level 3 or 4)
  and work up to a higher level (Total Aircraft).

23
Material Costs Studied

• Raw Material/Purchased Parts (RMPP)
• aluminum, steel, titanium, composites
• nuts, bolts, valves, hydraulic fittings, wires,
  cables
• Airframe structural parts made up primarily of
  RMPP
• Purchased Equipment (PE)
• manufactured and assembled items, procured from
  outside sources by the prime contractor
• typically higher in dollar value and more complex
  than a purchased part
• Subsystems (electrical, hydraulic,etc) fall under
  PE
• Avionics
• Radar, displays, communications, etc.

24
Material Cost Areas Evaluated

• Airframe RMPP /lb (AUW)
• Purchased Equipment /lb (AUW)
• Subsystems
• Landing Gear
• Avionics
• Aircraft RMPP PE /lb (AUW)
• Total Aircraft Procurement /lb (AUW)
• Additional costs (engineering and tooling
  recurring material ) included.

25
Normalization Process

• Cost data pulled directly from MACDAR Material
  Databases
• MACDAR provides unit cost data (Lot Average
  Cost), in raw TY.
• Escalated cost data to Constant Year FY01 using
  NAVAIR Commodity Indices.
• May obtain different results with OSD Indices
• Divided RMPP and PE Cost Data by AUW to obtain
  /lb.

26
Analysis Process

• Ran /lb against Aircraft Quantity Lot Midpoints.
  Included FMS quantities.
• Ran numerous iterations
• Ran each material category with and without EMD
• Ran Learn only
• Ran Learn with Rate
• Looked at Prime only material
• Looked at Prime Subcontractor material

27
Perform scatter plots
Enter data into COSTAT
Learning Curve Analysis
Learning and Rate Analysis
Evaluate stats plot Good?
Evaluate stats plot Good?
No
No
Disregard
Disregard
Yes
Yes
No
Evaluate Slope lt 100
Check learn rate results
Evaluate Slope Learn lt 100 and Rate lt 100
Investigate further
No
Yes
Yes
Publish Results
Publish Results
28
Results

• Airframe RMPP /lb
• obtained fairly good results when looking at one
  variant.
• some slopes in the mid to low 80s
• Purchased Equipment /lb and Avionics
• results not as good. Poor stats, possibly due to
• changes in configuration
• upgrades
• changes in vendor
• most slopes in the mid 90s
• Aircraft RMPP PE /lb
• obtained fairly good results
• some slopes in the mid to low 80s
• rate effect seen on one aircraft
• Total Aircraft Procurement /lb
• obtained fairly good results.
• some slopes in the mid to low 80s.

29
ResultsLearn only
30
ResultsLearn and Rate
31
ResultsSummary

• Steep slopes exhibited in some aircraft.
• Inconsistency seen among aircraft, leads to high
  uncertainty in material learning curves.
• Caution should be used when using slopes. May
  best be considered via risk analysis.
• Additional data points may decrease variability
  and uncertainty.

32
Future MACDAR Endeavors

• Increase scope and include newer platforms
• F/A-18 E/F, F-22, V-22
• Cargo and bomber platforms
• Modification programs

33
Back-up Slides
34
Escalation
35
Escalation Indices

• Number of different indices available
• OSD indices
• most widely known and accepted
• Broad based.
• Mandated by OSD for budgets.
• DRI WEFA maintains one of the largest privately
  available collections of economic and financial
  data.
• NAVAIR developed their own indices
• Utilize DRI Indexes more specific to Military
  Aircraft and its commodities.
• MACDAR uses the Fixed Wing Airframe Material
  Commodity Index and the Fixed Wing Electronic
  Material Commodity Index

36
Escalation Comparison OSD APN vs. Fixed Wing
Airframe Material
37
Escalation Comparison OSD APN vs. Fixed Wing
Airframe Material
38
Escalation Comparison Impact on Learning Curve
Slopes
39
Results Graphs
40
Results
41
Results
42
Results
43
Results
44
Results