Search papers, labs, and topics across Lattice.
This paper introduces the Graph-Structured Experiential Memory (GSEM) framework, which enhances multi-agent coordination in dynamic manufacturing by encoding historical coordination episodes as heterogeneous relational graphs. By leveraging a graph neural network for retrieval of structurally similar past experiences, GSEM enables agents to adapt their policies efficiently in response to operational disturbances. Experimental results demonstrate that GSEM significantly reduces makespan and adaptation time, particularly under conditions of high disturbance frequency, outperforming existing memory-augmented approaches.
Experience-guided policy adaptation can cut adaptation time by up to 38% in dynamic manufacturing environments.
Dynamic manufacturing environments require multi-agent systems to coordinate effectively under frequent operational disturbances such as machine failures, urgent job arrivals, and processing time variations. Existing multi-agent reinforcement learning approaches treat each disturbance episode independently, discarding valuable coordination experience that could accelerate future adaptation. In this paper, we propose a Graph-Structured Experiential Memory (GSEM) framework for multi-agent coordination in dynamic manufacturing. The framework encodes historical coordination episodes as heterogeneous relational graphs that capture task dependencies, machine states, and inter-agent collaboration patterns. When a new disturbance occurs, a graph neural network-based retrieval mechanism identifies structurally similar past episodes, enabling experience-guided policy adaptation rather than learning from scratch. Experiments on dynamic flexible job-shop scheduling benchmarks with three disturbance types show that GSEM reduces makespan by 4.1%-10.0% and adaptation time by 33%-38% compared to the strongest memory-augmented baseline, with the advantage increasing under higher disturbance frequency. Ablation studies and cross-disturbance transfer experiments further validate the necessity of graph-structured encoding and similarity-based retrieval and demonstrate the cross-disturbance generalizability of learned coordination patterns.
