Multi-Agent Systems for Supply Chain

Multi-agent systems (MAS) offer a natural paradigm for modelling, simulating, and operating modern supply chains. Each firm, facility, or functional role—supplier, manufacturer, distributor, carrier, retailer—can be represented as an autonomous software agent with its own goals, information, and decision policies. Agents interact through structured protocols to negotiate contracts, share forecasts, coordinate shipments, and resolve disruptions, mirroring the decentralised reality of cross-organisational supply chains.

Unlike monolithic optimisation approaches, MAS embrace distributed decision-making and partial information. This makes them well-suited to supply chain settings where no single entity has global visibility, participants have conflicting objectives, and local adaptation is essential for resilience. Recent advances in multi-agent reinforcement learning (MARL), mechanism design, and agent communication—particularly the integration of large language models as reasoning backbones—are significantly expanding what MAS can achieve in supply chain applications.

Key research areas include cooperative and competitive learning in multi-agent environments, contract net protocols and auction-based coordination, emergent behaviour in supply networks, and hybrid architectures that combine symbolic reasoning with learned policies. Contributions from the Supply Chain AI Lab at the University of Cambridge have demonstrated practical implementations of MAS for autonomous supply chain operations, bridging theory and industrial practice.

We invite you to explore the curated collection of key publications below, offering a gateway into this foundational and fast-evolving field.

List of Publications

1. Xu, L., Mak, S., Minaricova, M. and Brintrup, A., 2024. On implementing autonomous supply chains: A multi-agent system approach. Computers in Industry, 161, p.104120. [PDF]
2. Fox, M.S., Barbuceanu, M. and Teigen, R., 2000. Agent-oriented supply-chain management. International Journal of Flexible Manufacturing Systems, 12(2-3), pp.165-188. [PDF]
3. Swaminathan, J.M., Smith, S.F. and Sadeh, N.M., 1998. Modeling supply chain dynamics: A multiagent approach. Decision Sciences, 29(3), pp.607-632. [PDF]
4. Wooldridge, M., 2009. An introduction to multiagent systems. 2nd ed. John Wiley & Sons. [PDF]
5. Zhang, K., Yang, Z. and Başar, T., 2021. Multi-agent reinforcement learning: A selective overview of theories and algorithms. Handbook of Reinforcement Learning and Control, pp.321-384. [PDF]
6. Xu, L., Almahri, S., Mak, S. and Brintrup, A., 2024. Multi-agent systems and foundation models enable autonomous supply chains: Opportunities and challenges. IFAC-PapersOnLine, 58(19), pp.795-800. [PDF]
7. Monostori, L., Váncza, J. and Kumara, S.R.T., 2006. Agent-based systems for manufacturing. CIRP Annals, 55(2), pp.697-720. [PDF]
8. Mao, H., Zhang, Z., Xiao, Z., Gong, Z. and Ni, Y., 2020. Learning multi-agent communication with double attentional deep reinforcement learning. Autonomous Agents and Multi-Agent Systems, 34(1), p.32. [PDF]