A scenario based approach for flexible resource loading under uncertainty

Order acceptance decisions in manufacture-to-order environments are often made based on incomplete or uncertain information. To promise reliable due dates and to manage resource capacity adequately, resource capacity loading is an indispensable supporting tool. We propose a scenario based approach for resource loading under uncertainty that minimises the expected costs. The approach uses an MILP to find a plan that has minimum expected costs over all relevant scenarios. We propose an exact and a heuristic solution approach to solve this MILP. A disadvantage of this approach is that the MILP may become too large to solve in reasonable time. We therefore propose another approach that uses an MILP with a sample of all scenarios. We use the same exact and heuristic methods to solve this MILP. Computational experiments show that, especially for instances with much slack, solutions obtained with deterministic techniques for a expected scenario can be improved with respect to their expected costs. We also show that for large instances the heuristic outperforms the exact approach given a computation time as a stopping criterion.