DScheLLM: Enabling Dynamic Scheduling through a Fine-Tuned Dual-System Large language Model

2026-01-14

AI Summary

This study addresses the challenge of dynamic job shop scheduling under disturbances such as processing time fluctuations, machine failures, and unexpected tasks, where traditional methods suffer from limited generalization due to their reliance on specific event models. To overcome this limitation, the work introduces large language models (LLMs) into the domain for the first time, proposing a dual-system reasoning architecture that integrates “fast thinking” and “slow thinking.” The fast system efficiently generates high-quality scheduling solutions, while the slow system produces well-structured, solver-compatible decision inputs suitable for operations research solvers. The approach leverages exact data generated by optimization solvers for training and fine-tunes Huawei’s OpenPangu Embedded-7B model using LoRA. Experimental results on standard benchmarks demonstrate the method’s strong adaptability and practicality in handling previously unseen disturbances.

Technology Category

• Planning, Routing, and Scheduling: Planning with Language Models
• Reasoning under Uncertainty: Stochastic Optimization
• Machine Learning: Large Multimodal Models (LMMs)

Application Category

• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Production scheduling is highly susceptible to dynamic disruptions, such as variations in processing times, machine availability, and unexpected task insertions. Conventional approaches typically rely on event-specific models and explicit analytical formulations, which limits their adaptability and generalization across previously unseen disturbances. To overcome these limitations, this paper proposes DScheLLM, a dynamic scheduling approach that leverages fine-tuned large language models within a dual-system (fast-slow) reasoning architecture to address disturbances of different scales. A unified large language model-based framework is constructed to handle dynamic events, where training datasets for both fast and slow reasoning modes are generated using exact schedules obtained from an operations research solver. The Huawei OpenPangu Embedded-7B model is subsequently fine-tuned under the hybrid reasoning paradigms using LoRA. Experimental evaluations on standard job shop scheduling benchmarks demonstrate that the fast-thinking mode can efficiently generate high-quality schedules and the slow-thinking mode can produce solver-compatible and well-formatted decision inputs. To the best of our knowledge, this work represents one of the earliest studies applying large language models to job shop scheduling in dynamic environments, highlighting their considerable potential for intelligent and adaptive scheduling optimization.

Problem

• dynamic scheduling
• job shop scheduling
• production disruptions
• adaptability
• generalization

Innovation

• Dynamic Scheduling
• Large Language Model
• Dual-System Reasoning
• Fine-tuning
• Job Shop Scheduling