View article

We consider the mid-term production planning problem in make-to-order production, which consists in scheduling production orders so that the variability in the resource utilization over time is minimized. The problem is modeled as a resource leveling project scheduling problem, where production orders for final products and main components correspond to activities of a project. We present a new population-based resource leveling procedure of the iterated greedy type. The results of an experimental performance analysis including projects with up to 1000 activities indicate that the new method outperforms state-of-the-art resource leveling heuristics from literature.