A new heuristic to solve RCPSP with multiple execution modes and Multi-Skilled Labor

This paper examines a real production scheduling problem encountered in an assembly lines company. This problem generalizes the classic non-preemptive resource-constrained scheduling problem (RCPSP) in which a set of activities with a specific skill requirements has to be processed on a particular work center with limited capacity (cumulative resource). Activity processing times are often a function of the amount of labor with the appropriate skill allocated to its execution and the assignment has to be made in a pool of unary disjunctive resource (labor) associated with that skill. The objective function considered is minimization of the makespan. A lower bound is presented and a heuristic approach to solve this problem using the concept of dominance is proposed. The idea is to enumerate some schedulable combinations of activities with an appropriate resource assignment and schedule from them the one having the best value for an evaluation criterion. Finally, we compare our approach with some other classic, powerful heuristics based on priority rules and show that our results are competitive