Data for Costing Data for Parametric Costing Mort

Data for Costing Data for Parametric Costing Mort Anvari Director, Acquisition Costing (SAFM-CEA) 2008 Cost Research Workshop Acquisition Cost and Earned Value Reports: Data Quality Issue May 22, 2008

Key Points Data v Recognize the DCARC Data Collection Enforcement through Defense Acquisition Executive Summary (DAES) and the Success in Collecting Data for Costing at DCARC, Services, and PMs. v We have Several Data Quality Issues and Potential for Type I (α) and II (β) Errors in our Cost Estimating. Ø Examples include: STRYKER, FCS, and GFEBS. v Opportunity for Cost Analysis Enterprise Resource Planning (CA-ERP) and Data for Costing Architecture. Need Data for Costing: Future Technologies; Processes; Performances; Products; Services; Personnel; and Schedule. 2 May 22, 2008

Army DAES Group A CSDR Issues & Actions May 2008 Data 3 May 22, 2008

DAES Status Army Programs Data 6 11 11 DAES Status as of May 2008. Green: No or minor open compliance issues. Yellow : Unresolved minor open compliance issues. . Red: Any major issue. CSDR requirements not included on contracts. Unaddressed previously identified compliance issues. 4 May 22, 2008

Available Data for Costing Data v Cost and Software Data Reports (CSDRs) v Cost Performance Report (CPR) v Automated Cost Databases (ACDB) from CSDRs and CPRs v Contractor Logistic Support (CLS) Database v Cost Analysis Requirements Description (CARD) Database v Army Operating and Support Management Information System (OSMIS) v Army Capability Knowledge Base (CKB) v Additional data sources include: Ø Army Cost Positions (ACP) Database ACE-IT Sessions; SAR; DAES; WSR; APB; P/R-Forms; and Contract Summaries. 5 May 22, 2008

Data for Cost Estimating Data v Historical Data: Ø Cost and (Physical, Technical, Performance, Capability, Schedule). Information Knowledge Cost Models v Subject System Data: Ø Physical, Technical, Performance, Capability, Schedule. Performance Cost Schedule Estimates are only as good as the data that they are based on. 6 May 22, 2008

Data Issues Data v Current Data Reporting and Usage Issues (CSDR CPR). Ø Examples include: Stryker, FCS, and GFEBS (lack of clean data / no data). § Family or System of Systems § LSI, JV, Prime, Sub § CPIF, FFP, . . Reporting Requirements § Manufacturing, Services, SW, and ERP § Mod, ECP, and Incremental Contracts § Competition and Business Base Issues § Physical, Technical, TRL, Performance q CCDR, SRDR, and (Commonality Issue) (Contracting Issue) (Contract Type Issue) (Activity Type Issue) (Completeness Issue) (Quantity issue) (Technology Issue) KPP and Capability Data Issue Allocation Issue: Mapping to Standard Reports Labor – Material Direct – Indirect Recurring – Non-Recurring 7 May 22, 2008

Data Quality Challenges Data v Lack of data. Ø Physical, technical, performance, capability. Ø S/W ERP FFP Contract. Ø ACAT II and below programs. v Policies and procedures. Ø Do. DI 5000 -2 dated May 2003. § Requires SRDR but not CSDR for some programs. § Vague in some areas; CSDR “not required for procurement of commercial systems, or for non-commercial systems bought under competitively awarded, firm fixed-price contracts, as long as competitive conditions continue to exist”. Ø Army regulations do not require cost reporting. v Data collection process. Ø Existing data is not readily available to the Cost Analysts; sometimes they are not even aware that data exists. (Data in Jail) Ø Contractors required to map CSDR reports IAW Do. D MIL-HDBK-881. 8 May 22, 2008

Data Collection Process Contractor Data / Reports: R 3 Contractor 1 Contractor 2 Contractor Accountin g Systems: S 2 Contractor 3 S 9 S 7 S 1 S 8 S 6 S 3 S 4 S 5 S 10 R 11 R 10 R 18 R 14 R 17 R 8 R 12 R 5 R 7 R 16 R 2 R 4 R 13 R 15 R 9 R 6 Contractor Allocation Data Process vs. Quality. Labor vs. Material Direct vs. Indirect Rec. vs. Non Rec. Unique vs. Common Contractor allocation Contracting G 1 G 2 Data Quality? G 3 CSDR v Process affects data quality. Ø Increases probability for data errors. Ø Results in mapping / allocation inconsistencies. 9 May 22, 2008

Data Collection Process Data Contractor Data Input Form Contractor Database Contractor Allocation DCARC Database CSDR Army CES Allocation ACDB Database Contractor Accounting System DCARC / Do. D MIL-HDBK-881 ACDB / Army CES Sys 1 HDBK Element 1 CES Element 1 Sys 2 HDBK Element 2 CES Element 2 Sys 3 HDBK Element 3 CES Element 3 … … … Current Data Collection Process. v Contractor’s map data for CSDR reports to IAW Do. D MIL-HDBK-881. v CSDR data is then mapped into Army ACDB IAW Army Cost Element Structure (CES). 10 May 22, 2008

Automated Cost Database (ACDB) Data Army Automated Cost Database (ACDB) DCARC Database Collect Data of Historical Programs S Normalize CE io/n y Standardize m at Data Ar lloc A Input Data into ACDB Database Structure Cost Methodology Development Query ACDB for Cost and Technical Data Relate Technical and Program Parameters to Cost Perform Regression Analysis Develop Cost Models as appropriate ACDB. v Contractor data mapped twice. 11 May 22, 2008

FOUO STRYKER Data v Unit cost should not increase over time; suggests no learning. 12 FOUO May 22, 2008

FOUO STRYKER Data v Hull / Structure thought to be common but data does not support commonality. v Percentage of common parts for all HW WBS elements range from 11 - 39%. 13 FOUO May 22, 2008

FOUO STRYKER Data v Large variation in the unit costs for delivery orders 2, 10, 19, and 21. 14 FOUO May 22, 2008

STRYKER Data v TACOM review of Stryker CSDR. Ø Large variations in platform material costs across delivery orders. Ø Data suggested no learning. Ø Many HW elements %'s differ greatly between RDTE and Procurement. ; Ex: ICV; RDTE Hull/Structure ~9% of total HW cost; procurement ~37%. Ø Some cost elements were not in the category an analyst would expect; Ex: Tech pubs; operator manuals DO#17 and tech manuals DO#12. v Contractor’s accounting system collects data by delivery order. Ø Contractor MI-HDBK-881 reporting level did not provide insight into unique vs. common. (11 -39% common) Ø To conform to MI-HDBK-881 contractor allocated labor and some material costs. 15 May 22, 2008

STRYKER Data Shows commonality for variants. Commonality across all Stryker variants. Notional Resource Distribution Table (RDT). v CSDRs provide no insight into common versus unique. v Contractor collects data by delivery order and allocates labor and some material costs. 16 May 22, 2008

Future Combat Systems – FCS Data Prime Contractor Typical Program Sub-Contractors Boeing & SAIC FCS Prime Contractors (Lead System Integrators (LSI) Sub-Contractors FCS Family of Systems (Systems of Systems Approach). v Common development may be included in LSI data. Ø Examples include: SE/PM; data; etc. v Common versus unique. v Subcontractors have functional responsibilities across multiple platforms. 17 May 22, 2008

FCS Data Lead System Integrated (LSI) Tier I ANS MGVs Subsystems UAV MULE UGV Ground Sensor Integrated Computers Subsystems UGS Aerial Sensor Subsystems Tier II Mule UGV Functional Responsibility. v Tier I systems with functional responsibilities for MULE UGV: Ø LSI, Integrated Computers, Ground Sensor, and ANS. 18 May 22, 2008

FCS Software SLOC Data FCS includes 14 Integrated Weapon Systems and the Network • These Systems & Network could be a Separate Major Software Development Effort/Program System of Systems Software Microsoft Windows Vista: 50 Current FCS SLOC 97. 6 M SLOC on the Rise. Requirements too? 19 Total k. SLOC C 4 ISR 41, 672. 2 System of Systems Common Operating Environment (SOSCOE) 29, 866. 0 NSIV 1, 070. 7 IS&T 7, 889. 7 MGV 3, 649. 0 Logistics 1, 337. 3 Training 2, 201. 1 UAV 658. 2 UGV 9, 245. 8 Total FCS 97, 590. 1 May 22, 2008

Cost Data Flow Data Contractor Sys 1 Contractor Sys 2 ACDB Contractor Sys 3 OSMIS Contractor Sys 4 CLS Contractor Cost Accounting Systems Government Informational Systems / Databases Cost Analysts Current Information Access Flow. v Need for Data Warehousing. Ø Repository of electronically stored data. Ø Several databases typically supply data to the warehouse. Ø Improved ease of data retrieval and data quality. 20 May 22, 2008

Data Warehousing Data External Data Contractor Cost Accounting Systems Data Information Warehouse Delivery System Data Marts External Users Government Informational Systems / Databases Proposed Information Access Flow. v Raw Data loaded directly from contractor accounting system into data warehouse. v Metadata: ‘Data about Data’. v Data visualization / data warehousing / intelligent data mining. v Global definitions that can be referenced by many different databases. 21 May 22, 2008

Cost Analysis ERP Data v Potential for Enterprise Resource Planning (ERP) of Cost Analysis Data and Models across the Do. D. Ø Enable data and information sharing. Ø Allow automation and integration of business processes. v Data for Costing – Vision: Ø Collect raw data with more flexibility and focus on metadata development – XML. Ø Data warehousing of raw and current data. Ø Acquiring data mining and search tools to assist data analysis. Cost Analysts need to be more involved in the process. Why Data ERP for Costing? Check out http: //www. anvari. net/E_Proc_Model. htm 22 May 22, 2008