Efficient Order Batching Optimization Using Seed Heuristics and the Metropolis Algorithm

Abstract Order Picking in warehouses is often optimized using a method known as Batching, which means that one vehicle can be assigned to pick batch of several orders at time. There exists rich body research on Batching Problem (OBP) optimization, but area demands more attention computational efficiency, especially for optimization scenarios where have unconventional layouts and capacity configurations. Due the NP-hard nature OBP, cost optimally solving large instances prohibitive. In this paper, we compare performance two approximate optimizers designed maximum efficiency. The first optimizer, Single Batch Iterated (SBI), based Seed Algorithm, second, Metropolis Sampling (MBS), algorithm. Trade-offs memory CPU-usage generalizability both algorithms analyzed discussed. Existing benchmark datasets are used evaluate various scenarios. On smaller instances, find come within few percentage points optimality minimal CPU-time. For larger solution improvement continues throughout allotted time rate difficult justify many operational SBI generally outperforms MBS mainly attributed search space latter’s failure efficiently cover it. relevance results Industry 4.0 era warehouse operations