A Sliding-Window-Based Reinforcement Learning for Dynamic Assembly Flow Shop Scheduling with Multi-Product Delivery
2026-07-03 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors studied how to better schedule tasks in a factory system where many products need to be assembled and delivered together, which is tricky because orders come in randomly and affect how jobs and machines are assigned. They created a new method using reinforcement learning that looks at a small time window of tasks and uses a special graph model to understand complex dependencies and bottlenecks. Their approach prioritizes important steps, adapts to changing conditions, and was tested on real factory data, showing improvements over traditional scheduling methods and other AI approaches. The results indicate their method is effective and flexible under different production scenarios.
reinforcement learningschedulinghybrid manufacturing systemskitting deliveryMarkov decision processgraph neural networksbottleneck dynamicsdispatching rulesflow shop schedulingreal-time scheduling
Authors
Junhao Qiu, Jianjun Liu, Ting Liu, Rongjie Liao, Zhantao Li, Qingfu Zhang
Abstract
Multi-product kitting delivery imposes significant challenges for real-time scheduling in hybrid manufacturing systems that integrate processing and assembly, as dynamic order arrivals simultaneously alter supply dependencies and the set of feasible job-machine assignments. This paper proposes a sliding-window-based reinforcement learning (SWRL) framework for end-to-end online scheduling in the flexible assembly flow shop scheduling problem with complex kitting constraints. The problem is formulated as a heterogeneous graph-based Markov decision process that captures the dual-layer kitting structure and the tail-product bottleneck dynamics that produce a sparse reward landscape. To address the resulting challenges, SWRL integrates a sliding-window filtering mechanism that filters inactive nodes and prioritizes kitting-critical operations, a spatiotemporal graph encoding network that tracks bottleneck shifts across consecutive decision states, and a dynamic action mapping module with a constrained waiting strategy that adapts to the changing action space under variable topologies. Experiments on real-world instances from a home appliance manufacturer demonstrate that SWRL achieves consistent tardiness reductions over classical dispatching rules and existing deep reinforcement learning methods, and exhibits robust performance across varying resource configurations, order loads, and arrival concentrations.