GAs and simulation techniques for dynamic resources sharing and reallocation across workgroups

This work proposes a decision support system (DSS) for the management of resources scheduling emergencies in labor-intensive industrial and business contexts. Prior work has demonstrated the benefits that may be accrued from sharing and re-assigning resources across workgroups in case of unexpected manpower shortage in specific areas of competence. The decision process to identify suitable substitute resources that best fit the allocation criteria specified for the job/task within the pool of existing resources is quite complex because of the multiplicity of interdependent factors that concurrently influence performance. These include: cost, quality of service, and service completion time. The process becomes even more complicated in work environments characterized by multiple resources with quite diversified skills and competencies. The DSS proposed in the paper integrates artificial intelligence (AI) techniques based on genetic algorithms (GAs) and simulation to iteratively generate and test possible reallocation scenarios until an optimal trade off among the performance drivers is found. The paper illustrates the key features of this hybrid DSS and presents the results from a demonstrator developed for a prototype retail store.