In the consumer packaged goods industry, usually a large number of final items are produced in a highly automated flow line production system. Florian Seeanner presents a mathematical model which is an improved version of the General Lot-Sizing and Scheduling Problem for Multiple production Stages (GLSPMS) to tackle the problem of shifting bottlenecks. Especially, a new meta-heuristic is developed which combines the principles of Variable Neighborhood Decomposition Search (VNDS) and Exchange. This VNDS+E heuristic is able to find very good solutions for the improved GLSPMS, including a real-world problem case. Finally, further model extensions, e.g., for scarce setup resources, are proposed.