Collaborative Human–Robot Teaming for Dynamic Order Picking: Interventionist strategies for improving warehouse intralogistics operations

Order picking, a labor-intensive and costly aspect of picker-to-parts warehouse systems, faces challenges such as rising worker injuries, faster delivery expectations, and dynamic order arrivals. Collaborative automation, particularly human-autonomous mobile robot (AMR) collaboration, offers a promising solution, but limited research addresses its role in managing dynamic orders. This study introduces the Collaborative Human–Robot Dynamic Order Picking Problem (CHR-DOPP), where workers perform pick-and-place tasks while AMRs handle transportation. We propose an interventionist strategy to manage dynamic orders by allowing ongoing AMR and worker pick cycles to be updated with new requests. Two interventionist algorithms are proposed: a reactive strategy that dynamically assigns new orders based on AMR availability and proximity, and a conditional strategy that selectively integrates new requests to minimize disruption to ongoing pick cycles. Their performance is benchmarked against traditional human-only picking with intervention and non-interventionist collaborative strategies. The system performance is assessed using three key measures, namely, average order completion time (AOCT), average worker travel distance (AWTD) and average total tardiness (ATT). Extensive numerical experiments are conducted to assess the performance of the proposed interventionist algorithms, including impact of routing strategy, AMR cart capacity, number of warehouse zones, and nature of order arrivals. Our results show that the proposed strategies outperform the traditional human-only picking strategy across all three metrics (AOCT, ATT, and AWTD), while achieving significantly better AOCT and ATT compared to a collaborative system without interventions, despite an increase in AWTD. Finally, numerous managerial insights for different warehouse types are provided based on our findings.