Energy-Efficient Shift Staffing and Machine Stopping: A Threshold-Driven Two-Level Heuristic

This paper seeks to introduce an energy-aware heuristic for smart manufacturing. The core of the heuristic is a dispatching rule which supports two different types (levels) of decisions, namely (1) whether night-shifts should be staffed or not and (2) whether a particular machine should be stopped or not. While existing approaches often focus on energy-aware machine control, they typically neglect the more expensive cost implications of night-shift personnel availability. This paper addresses that gap by combining both machine- and system-level decisions into a unified heuristic. The decision is based on the current (day-time-dependent) energy price and the current workload within the production system, using predefined thresholds for either criterion. To evaluate the proposed heuristic, it is embedded into a stochastic multi-item, multi-stage simulation model of a manufacturing system. Cost aspects include energy, production logistics, and labor costs for staffing the night shifts. Computational results confirm that the trade-off between those cost factors can be effectively controlled via the parameterization of the underlying dispatching rule, i.e., the concrete threshold values. Furthermore, the lowest overall costs can be achieved when night-shifts are enabled only in case of very low energy prices. On the lower (machine) level, it turns out that allowing machines to operate also under moderate energy prices or small workloads helps avoid pushing too much work content towards the night-shifts.