Autonomous Scheduling and Dynamic Resource Allocation in Smart Manufacturing Using Multi-Agent Systems: A Comprehensive Review

Abstract

The rise of customizable, high-mix low-volume production schedules withinmodern smart manufacturing has exposed significant limitations in traditional centralized production scheduling systems. Conventional planning methods, which typically rely on monolithic enterprise resource planning (ERP) systems and static centralized dispatch heuristics, lack the responsiveness required to manage real-time shop-floor updates. These updates include frequent machine breakdowns, expedited emergency orders, material processing delays, and changing tool availability. To transition toward highly adaptable smart factories, Multi-Agent Systems (MAS) have emerged as a leading decentralized architecture for achieving autonomous scheduling and dynamic resource allocation. This review paper provides a comprehensive, state-of-the-art analysis of MAS architectures deployed in manufacturing environments. We systematically evaluate the structural design principles of MAS, investigating alternative auction-based protocols, Contract Net Protocols (CNP), and collaborative negotiation mechanisms that allow independent agents to distribute manufacturing tasks autonomously. The paper details how physical shop-floor elements—such as workpieces, individual machine tools, automated guided vehicles (AGVs), and storage areas—are mapped into specialized autonomous software entities. Furthermore, we identify key research gaps in current literature, including deadlocks in complex multi-agent negotiations, scalability challenges under highly distributed multi-facility networks, and the lack of standardization for multi-protocol communications. Finally, a structured development roadmap is proposed to guide future research toward integrating deep reinforcement learning with multi-agent coordination loops, establishing robust, self-optimizing. KEYWORDS: Smart Manufacturing; Multi-Agent Systems; AutonomousScheduling; Dynamic Resource Allocation; Industry 4.0; Contract NetProtocol.