A Review of Research on Multi-Agent Technology for Solving Collaborative Scheduling Problems

The increasing complexity and dynamics of modern industrial environments challenge traditional centralized scheduling methods, necessitating more adaptive and intelligent approaches. This review systematically examines the application of multi-agent system (MAS) to collaborative scheduling problems in industrial production. A systematic literature review was conducted, covering 51 peer-reviewed articles published between 2017 and 2026 (databases searched on 31 March 2026), drawn from major databases. This study adopts the stimulus-organism-response (S-O-R) framework to synthesize and structure the fragmented literature, offering a novel theoretical lens that links environmental stimuli, agent internal mechanisms, and scheduling performance. Key challenges are identified, including task allocation trade-offs, dynamic rescheduling conflicts, and human-robot collaboration within Industry 5.0. The analysis reveals that while MAS significantly enhances flexibility, robustness, and responsiveness compared with centralized approaches, issues such as communication latency, decision volatility, and scalability persist. The proposed S-O-R framework integrates cognitive mechanisms and interactive behaviors to explain how agent intelligence mediates scheduling performance. This review concludes by outlining future research directions and emphasizing the need for real-world validation, consideration of ethical implications, and alignment with human-centric, resilient, and sustainable manufacturing values.