Stigmergic Graph Memory: An Environment-Aware Approach for Many-to-Many Multi-Agent Pickup and Delivery
Multiagent SystemsRobotics
Summary
The authors studied how robots in a busy warehouse pick up and deliver items without causing traffic jams. Instead of just planning routes after deciding where to go, they made a system called Stigmergic Graph Memory (SGM) that remembers recent robot movements to help choose better pickup and drop-off points. This memory helps the warehouse run more smoothly by suggesting goals that avoid congested spots. Tests showed that their method improved the number of tasks completed by about 20-37% compared to older methods.
Authors
Aditya Dutta, Joon-Seok Kim
Abstract
Automated fulfillment warehouses must continuously assign and execute pickup-and-delivery work while avoiding congestion. In many-to-many Multi-Agent Pickup and Delivery (MAPD), a request specifies a stock-keeping unit rather than fixed endpoints, requiring the controller to select an agent, source, and destination before path planning. Existing graph-guidance methods primarily influence routing after goals are fixed, leaving endpoint instantiation uninformed by recent traffic. We introduce Stigmergic Graph Memory (SGM), a bounded, decaying memory layer that records recent execution signals on warehouse nodes and directed edges to rank feasible endpoints and route preferences without altering collision constraints or planner validity. Across paired request streams on five layouts, three load levels, and 25 seeds per condition, SGM outperforms two reconstructed many-to-many allocation baselines in all 15 map-load conditions, with paired throughput gains of 20.5-36.7%. These results show that recent execution memory can improve warehouse throughput by shaping which feasible goals enter the planner, not only how agents travel to already fixed goals.