Improving human–robot collaboration in AMR-assisted warehouse picking

Abstract In AMR-assisted picker-to-parts warehouses, performance loss arises not only from inefficient robot tours but also from poor synchronization between AMR arrivals and picker readiness at handover points. To address these problems, prior research has largely emphasized batching, routing, and sequencing, with limited attention to handover-level online synchronization. Focusing on this collaborative control gap, the research reported in this article develops a HUB-based Collaborative Intelligence (HUB-CI) coordination layer and a State-aware Human–Robot Synchronization (SHRS) protocol for zone-based collaborative picking. Both are based on collaborative control theory. HUB-CI harmonizes distributed picker and AMR execution states and converts them into prediction-ready synchronization variables. SHRS then uses these variables to perform event-triggered next-handover dispatch, minimizing the synchronization mismatch at the handover point. The framework is evaluated through a discrete-event simulation using mapped layout geometry and order data from a footwear manufacturing warehouse management system. Relative to a cyclic tour benchmark, SHRS reduces mismatch by 42% and improves throughput by 36%; relative to a nearest-neighbor dispatching benchmark, the corresponding improvements are 16% and 9%. These improvements arise from SHRS explicitly minimizing predicted human picker-AMR arrival mismatches at each dispatch epoch, rather than relying solely on fixed tour order or travel proximity. Additional disturbance, sensitivity, and statistical analyses show that SHRS generally retains its performance advantage across the tested operating conditions. Overall, the results indicate that, within the modeled warehouse configuration and operating conditions, handover synchronization is an important operational factor in collaborative picking performance, and that HUB-CI-enabled state-aware control provides a promising direction for real-time adaptive warehouse operation.