The impact of zoning policies on efficiency in collaborative picking systems with autonomous mobile robots

The increasing deployment of robotic technologies have transformed warehouse operations into highly complex collaborative environments. In picker-to-part systems, human pickers and automated mobile robots (AMRs) must coordinate effectively to ensure timely order fulfilment. Although optimisation models can generate efficient routing and batching plans, they are often difficult to implement in practice, as human operators may not consistently follow complex routing instructions. Consequently, many warehouse management systems adopt zoning-based policies to simplify picker tasks. However, the performance implications of such policies in collaborative human–robot settings remain largely underexplored.
This study investigates the impact of zoning decisions in a collaborative order-picking system where AMRs transport batches of orders and human pickers retrieve and load items onto the robots. The key decisions include batching orders for AMRs, assigning pickers to zones, and routing robots and pickers. We analyse three coordination strategies: no-zoning policy, static zoning policy with fixed picker areas, and a dynamic zoning policy that allows zones to change across picking cycles. We develop a mathematical model and a multi-stage heuristic for each zoning strategy. Computational experiments on various warehouse settings provide managerial insights into the effectiveness of zoning policies and highlight how dynamic zoning can balance practical simplicity and performance in collaborative human–robot order-picking systems.