Efficient algorithms for production scheduling

Scheduling problems play an important role in the area of production planning. However, due to e.g. uncertainties, real-world applications may induce additional constraints, and lead to intractable models. In the literature, approximated solutions are often computed. This thesis aims to derive exact yet tractable algorithms for different scheduling problems under robustness or inventory constraints. First, we consider the notoriously NP-hard Buffer Allocation Problem (BAP) in flow lines. In its classical approach, it assumes that the processing times of jobs are known in advance. Realistically, this is not the case. Therefore, we present a model for the BAP with additional robustness constraints. We compute exact solutions and demonstrate the tractability of our method. Next, we lay focus on inventory-constrained scheduling. In this setting, jobs are assumed to add or remove a given amount of material from a common stack. We identify a new class of such problems, where the objective function only depends on the consuming jobs. We provide complexity results and algorithms for variations of the problem with different objective functions and constraints.