Optimized staffing for product releases and its application at Chartwell Technology

• Authors: Puneet Kapur;An Ngo-The;Günther Ruhe;Andrew Smith
• Venue: Journal of Software Maintenance and Evolution: Research and Practice - Search Based Software Engineering [SBSE]
• Year: 2008

Release planning for incremental software development assignsfeatures to releases such that technical, resource, risk and budgetconstraints are met. Each feature offers a piece of functionality.A feature can be offered as part of a release only if all itsnecessary tasks are done before the given release date. These tasksrequire different skills. Staffing for product releases asconsidered in this paper is the process of assigning humanresources from a given pool of developers who might have varyinglevels of skill to perform different tasks. In addition to that, weconsider time windows of absence of the developers. The primarygoal of staffing is to provide product releases of best qualitywhere quality means offering the most attractive features tocustomers in a timely manner. We call the problem STAFF-PRO. Theproblem is known to be NP-complete. Consequently, we have to besatisfied with solutions that are sufficiently good, but notnecessarily optimal in the case of mid-sized or large problems.Search-based methods relying on meta-heuristics have been proven tobe successful in similar contexts. In this research, a focusedsearch (FS) method is presented. This refers to a two-phasedsolution approach where Phase 1 applies integer linear programmingto a relaxed version of the full problem. Its solution is used as astarting point to perform FS in a reduced search space in Phase 2.The search itself is conducted by a genetic algorithm. It generatesa solution that fulfills all the stated resource and schedulingconstraints and is of a proven degree of optimality. We performedan empirical analysis of the proposed solution approach bycomparing FS and unfocused search (UFS) (without Phase 1) for aseries of 200 test examples. On average, FS performs about 15%better than UFS. The whole method was applied as an industrial casestudy performed at Chartwell Technology. The case studydemonstrates that application of the FS method to STAFF-PRO (i)allows a reduction in the time needed for generating acceptablestaffing plans, (ii) generates plans of proven quality that arebetter than manual plans and (iii) supports the various types ofre-planning necessary for varying parameters, budgets and resource.Copyright © 2008 John Wiley & Sons, Ltd.