1. Copyright © 2012 by Gan Wang, Lori Saleski and BAE Systems. Published and used by INCOSE with permission. The Architecture and Design of a Corporate Engineering Data Repository Dr. Gan Wang Ms. Lori Saleski 27 th International COCOMO Forum 16-18 October 2012 Software Engineering Institute, Pittsburgh, PA

2. Outline  What is the Engineering Data Repository?  CONOPS  Architectural and Design Considerations  User Interfaces  Data Collection Process  Lessons Learned 2

3. What’s the Engineering Data Repository?  A suite of tools, comprised of –Database of current and historical program data –Data search and query portal –Estimating tools and process –Data collection and process –Analysis and calibration tools 3

4. What’s in the Repository? Historical Data:  Project metadata –Project and program level descriptions –Attributes and characteristics  Engineering financial actuals –Actual (historical) efforts –Estimated (in-execution) efforts  System-level design metrics, e.g., –Requirements –Interfaces –Algorithms –System complexities  Subsystem/component-level metrics, e.g., –SLOC counts –SWaP –Drawing counts –Supportability Tools and Processes:  Engineering (parametric) estimating tools  Data collection and analysis tools  Procedures, templates, checklists  Training materials, videos  References 4

5. What’s Motivation?  The dilemma we ran into in the past… 5 D ESIGN. Adapted from Dilbert, © Scott Adams, Inc.  Provide enterprise-wide, central and easy-to-use data store for all engineers –Consistent use of historical data in system design and business winning –Supporting entire project life cycle –Improved estimation accuracy, confidence and credibility –Save cost and reduce cycle time –Strategic competitive discriminator

6. Lifecycle Operational Data Needs 6 Use of historical data essential in system design and program execution

7. Typical Use Cases 7 Query Single Data Point Query Single- Attribute Statistics Query Group of Data Points Ingest Single Data Point Data Repository + + + Data Collector Data Analyst Everyone

8. Three Interconnecting Logical Databases 8 In-Progress Program Database Estimate Database Post Mortem Database Use Past Results to Predict Future Performance Business CaptureProgram ExecutionProcess Improvement Project Engineering EACsProject Engineering Actuals Cost Proposals Time Project Bid / Proposal Project Life CycleProject Post Mortem Engineering Estimates

9. Cost Element Structure 9 1.0 – System/Project 1.1 – Integrated Project Management 1.2 – Systems Engineering 1.3 – Prime Mission Product (PMP) 1.4 – Platform Integration 1.5 – System Test & Evaluation (ST&E) 1.6 – Training 1.7 – Data Management 1.8 – Peculiar Support Equipment 1.9 – Common Support Equipment 1.10 – Operational / Site Activation 1.11 – Industrial Facilities MIL-STD 881 Compliant Mapping Rules Project/Organization Specific Structures

10. Dynamic Data Structure Design 10 Dynamic linked lists used to support different data types and periodic collections throughout project and system life cycles

11. Database Workflow Supporting Collection of Multiple Data Instances of Same Program Data 11 Built-in Data Ingestion and Review Functions to Support Role-based Interactions Access Control by Roles

12. Data Query Portal – Search by Program Attributes 12 Search Criteria Accessible from Corp. Intranet Search Dashboard Search Relevance

13. Detail Data Queries 13 Calculated Statistics Detail Project Data

14. Consistent Process is a Cornerstone Is project complete? Clear start and finish? Clear funding line? Is it relevant to future bids? Project points of contact found? Project financial, technical and other supporting data available? 14 Qualification of Data Point Identify historical programs Quality program Establish program technical and financial POCs Conduct initial data collection training session Learn the Data Collection Template Learn metrics definitions Collect raw data Financial database Technical data records Interview and data review sessions Recollect questionable data Additional interview sessions Conduct analyses Normalization Statistical analysis

15. Practical Lessons Learned  Data quality –Consistency, consistency, consistency! –“Apples-to-oranges” has little to no benefit  Plan for additional time –Assumptions/definitions of data –Analysis in aggregate  Problem solving exercise  Corporate expertise  Management championship 15

16. In Closing  Enterprise-level Engineering Data Repository –Database of current and historical program data –Data search and query portal –Estimating tools and process –Data collection and process –Analysis and calibration tools  Used throughout project and system design life cycles  Proven effective –Easy-to-use, popular with engineers –Improves accuracy, confidence and credibility –Saves cost and reduce cycle time –Enhances business competitiveness 16

17. 17 Dr. Gan Wang gan.wang@baesystems.com 703-668-4259 Ms. Lori Saleski lori.saleski@baesystems.com 603-885-6353 Questions & Comments