Joint optimization of order- and rack-scheduling in KIVA picking systems

We study the order processing operations in KIVA robot-assisted warehouses, where racks are delivered to multiple workstations by robots so that pickers at each workstation just focus on retrieving items from the racks to fulfill the orders. In this context, the order- and rack-scheduling, including their assignment and sequencing decisions, should be considered integrally as they are closely related and can enhance systemic efficiency by leveraging their synergy. We thus consider the comprehensive problem of jointly allocating orders and racks to workstations under workload balancing and sequencing their interlinked processing flows. We formulate it as a mixed-integer programming model to minimize the total number of rack visits. To tackle this model, we present a simulated annealing search framework, which builds on a relaxation model and a best-first-search heuristic exploiting the problem structure. Computational studies show that our proposed approach performs well on small-sized instances. On a large scale, it outperforms both the rule-based policy and two other state-of-the-art algorithms in terms of solution quality. We also conduct sensitivity analysis to generate some managerial insights for real-world warehouse operations.