A Genetic Algorithm for Scheduling both Jobs Families and Skilled Workforce

Generally, solving a regular group scheduling problem consists of sequencing both jobs within each group and groups themselves, with the aim of optimizing a specific performance indicator. However, one cannot but notice that human resources are getting more and more critical in production systems, and it is not enough to concentrate only on the physical resources in order to expect a reasonable solution. This paper studies how the allocation of M differently skilled workers to machines may affect the makespan minimization for a M-machines flow-shop group scheduling problem with sequence dependent setup times. In particular, a team of heterogeneous workers have to be assigned to machines for executing setup tasks between one group and another, thus affecting the response of a work cell with a flow shop layout. A twofold objective characterizes the proposed research: 1) since the problem is NP-hard, finding the best heuristic optimization strategy to run both group/job sequencing and worker allocation problems; to this aim, three distinct genetic algorithm-based approaches have been tested on the basis of a properly designed benchmark of test cases. 2) Making full use of the best metaheuristics to evaluate the impact of the workers’ skills on both manpower cost and makespan.