23rd International Forum on COCOMO® and Systems/Software Cost Modeling and ICM Workshop 3
Theme: Cost and Schedule Estimation for Next-Generation Processes
Davidson Conference Center, USC - October 27-30, 2008
Our effort to date in developing guidance for using the Incremental Commitment Model as a next-generation process for developing and evolving future systems has identified a number of emerging challenges for systems and software cost estimation and associated management metrics. Future systems with emergent requirements will need to deal with a broader early Cone of Uncertainty, and will need increased investments in early uncertainty and risk reduction, such as Competitive Prototyping. The increasing importance of evidence-based reviews (vs. schedule-based or event-based reviews) as a basis of readiness to proceed increases the criticality of good planning, estimating, and earned value progress tracking of evidence development.
The increasing pace of change in threats, missions, technology, and the environment creates a forward-expanding Cone of Uncertainty that highlights the need for relatively short, stabilized development increments that reduce the risk of delivering obsolete or uncompetitive system releases. Continuing competitiveness also implies the need for further investments in analyzing sources of change and adapting future-increment plans and priorities while the stabilized team is developing the current increment. Estimating and managing the resources for this mode of development creates further challenges for current estimation models, progress metrics, contracting methods, and incentive structures. It also blurs the boundary between traditional systems and software "development" and "maintenance," and highlights the importance of accurately estimating an increasingly critical parameter that we are calling the Incremental Development Productivity Decline (IDPD) factor.
Many traditional incremental development projects (and entire subfields such as agile methods and its "yesterday's weather" productivity heuristic) have optimistically assumed that their productivity for later increments will be roughly the same as for earlier increments. If staffing is stable, there can be some learning-curve productivity improvements. But such factors as breakage and maintenance of the existing code base, diseconomies of scale in development and integration, requirements volatility, and user demand for enhancements to the existing code base will consume significant developer effort and generally far outweigh any next-increment productivity gains.
We will address these issues in the upcoming COCOMO®/SSCM Forum and Incremental Commitment Model (ICM) workshop. On Monday and Tuesday, October 27-28, 2008, we will focus the Forum on cost and schedule estimation issues for next-generation systems and software engineering processes, including a tutorial and panel on the topic, several related paper sessions, and further panels on cost model and tool integration and other future trends in systems and software estimation methods. On Wednesday and Thursday, we will have one of our continuing COSYSMO workshops on aspects of COSYSMO 2.0 (Ricardo Valerdi’s COSYSMO 1.x book is almost all at the publisher), plus a continuation of our series of workshops on the guidebook for using the Incremental Commitment Model on DoD projects. This will include updates on our draft guidebook work to date and analysis of industry Competitive Prototyping (CP) survey responses, complementary industry and government presentations, and working group sessions on the software aspects of CP and on next-generation estimation methods and management metrics for next-generation processes such as the ICM.
General Chair: Barry Boehm, USC Program
Committee: Rick Selby (Northrop Grumman/USC, Chair), Barry Boehm (USC), Winsor Brown (USC), Sunita Chulani (IBM), Ed Colbert (USC), Jared Fortune (USC), LiGuo Huang (SMU/USC), Jo Ann Lane (USC), Dan Ligett (Softstar), Ray Madachy (USC), Ali Malik (USC), Chris Miller (OSD), Vu Nguyen (USC), Don Reifer (RCI/USC), Thomas Tan (USC), Ricardo Valerdi (MIT/USC), Ye Yang (ISCAS/USC)
Local Arrangements Chair: Julie Sanchez, USC (email@example.com)
|Key Dates:||Forum and Workshops:|
|1-3 page Abstracts Due:
August 8, 2008 to firstname.lastname@example.org
|Oct. 27 – 28: Forum (Contributed presentations, Tool Fair); Affiliates’ Dinner|
September 5, 2008
|Oct. 29 – 30: Workshops (Incremental Commitment Model Guidebook III; COSYSMO 2.0; TBD)|
| Proceedings Materials Due:
October 10, 2008