real-time reoptimization

This work addresses the challenge of rapidly adapting deployed operations research optimization models to new constraints or disturbances in dynamic real-world environments, where current approaches heavily rely on expert intervention. We propose a novel large language model (LLM)-based agent framework that embeds an LLM as an operations research expert within the reoptimization pipeline. The framework leverages natural language interaction to interpret evolving requirements, automatically generates structured model patches, and incorporates information from prior solutions to design acceleration strategies, enabling efficient and interpretable continuous adjustment. By integrating valid inequalities, solver tuning, and metaheuristics, the method demonstrates strong empirical performance in both online supply chain reoptimization and offline university exam timetabling, significantly improving computational efficiency while preserving solution quality and enabling rapid response with minimal expert dependency.

Traditional offline construction of Multi-Schedule Graphs (MSGs) fails to accommodate dynamic scenarios such as hardware failures and runtime slack variations, resulting in insufficient robustness of AI-driven scheduling in time-triggered systems. This paper proposes an online adaptive scheduling method based on reinforcement learning (RL), embedding an RL agent within a meta-scheduler to enable real-time MSG expansion, continuous scheduling policy optimization, and support for context-aware adaptation, mode switching, and dynamic performance-bound adjustment. Our key contribution lies in overcoming the limitations of offline training by enabling incremental, runtime construction of the MSG and concurrent exploration of scheduling policies. Experimental evaluation under strict deadline constraints demonstrates a 23.6% improvement in scheduling success rate and enhanced fault recovery capability; moreover, response latency to timing fluctuations and unexpected events is reduced by 41%, significantly improving system reliability and adaptability.

Dynamic flexible job shop scheduling faces the challenge of simultaneously achieving millisecond-level real-time responsiveness and long-term global optimization. To address this, this work proposes RACE-Sched, a framework featuring an asynchronous dual-stream architecture that decouples execution from reasoning: a reactive stream employs low-latency symbolic heuristics for immediate scheduling decisions, while a deliberative stream leverages large language models to parallelly generate, validate, and evolve scheduling rules. The framework innovatively incorporates a semantic rule repository to enable cross-scale transferability and integrates sandbox validation with atomic update mechanisms to ensure system safety. Experimental results demonstrate that RACE-Sched significantly outperforms existing deep reinforcement learning and LLM-based approaches on GEN-Bench, MK-Bench, and JMS-Bench, achieving superior performance in both scheduling quality and dynamic adaptability.

HPC workloads are becoming increasingly heterogeneous, rendering traditional static heuristic schedulers inadequate for dynamic resource demands. To address this, we propose SchedTwin—the first real-time digital twin system for HPC job scheduling. It continuously ingests runtime event streams to drive high-fidelity discrete-event simulation, enabling rapid online evaluation of “what-if” scenarios across multiple scheduling policies and facilitating goal-driven, closed-loop adaptive scheduling. Deeply integrated with the PBS scheduler, SchedTwin achieves low-overhead (sub-10-second decision latency) and high-accuracy online policy optimization. Experimental evaluation in production environments demonstrates that SchedTwin significantly outperforms mainstream static schedulers—overcoming the longstanding dual bottlenecks of adaptability and timeliness inherent in conventional HPC scheduling approaches.

This study addresses the challenge of objectively evaluating algorithm performance in the Dynamic Flexible Job Shop Scheduling Problem (DFJSP), which is hindered by reliance on static benchmarks and uncalibrated instance generators. To overcome this, the authors propose DynaSchedBench, a diagnostic framework featuring a Sequential Event Space Calibrator (SESC) that computes a Scheduling Stress Index (SSI) to enable controllable generation of problem instances with tunable difficulty. The framework supports snapshot-based simulation, agent testing, and visualization. It achieves, for the first time, precise and efficient control over DFJSP instance hardness. Empirical analysis reveals an “observability paradox” in large language model (LLM)-based scheduling agents—access to complete information unexpectedly degrades performance. Furthermore, most LLM agents exhibit only heuristic-level approximation capabilities, failing to surpass strong handcrafted heuristics, with limited gains from tool augmentation.

This work proposes a novel paradigm that integrates discrete-event simulation with a large language model (Gemini-1.5-Pro) to overcome the limitations of traditional simulation-based optimization, which treats simulators as black boxes and offers little insight into policy failure. By leveraging event-level trajectory replay, the method automatically identifies bottlenecks from low-scoring simulation runs and generates interpretable, traceable, code-level policy revisions in parallel. It pioneers the use of simulation trajectories to guide the LLM in targeted heuristic rule modification, combined with rolling evaluation and an elite retention mechanism for iterative policy improvement. Evaluated on dynamic production and AGV scheduling tasks, the approach achieves an average policy score of 77.51 (out of 100), improving the best run from 62.49 to 78.61, and significantly outperforms MILP, handcrafted rules, and metaheuristic baselines across 100 random seeds and fault perturbations.

This work addresses the challenges in industrial scheduling where asynchronous event streams often lead to inconsistent decision states, ambiguous action validity, and difficulties in attributing execution errors in reinforcement learning policies. To resolve these issues, the paper proposes a policy-decoupled execution and measurement layer that bridges the policy and the execution environment. By constructing valid decision snapshots, defining standardized execution contracts, and recording multidimensional execution deviations, the approach structurally formalizes execution semantics for the first time. This enables observable and attributable deployment discrepancies between simulation and reality, transforming ambiguous execution failures into type-labeled supervisory signals. Experimental results demonstrate that the framework consistently enhances diagnostic capability across varying observation delays, significantly reducing avoidable errors under low-latency conditions and providing structured supervisory data for policy evaluation and optimization.

This work addresses the challenge of maximizing end-to-end success probability in structured agent workflows under hard constraints on budget and deadline. The authors propose Monte Carlo Combinatorial Planning (MCPP), a lightweight closed-loop planner that dynamically replans during execution in response to observations. MCPP employs a finite-horizon stochastic online allocation model with parallel sampling and leverages Monte Carlo simulation to estimate, in real time, the probability of successful task completion under the given constraints. Experimental results demonstrate that MCPP significantly outperforms strong baseline methods on the CodeFlow and ProofFlow benchmarks, consistently achieving higher task completion rates across diverse budget–deadline configurations. These findings validate MCPP’s effectiveness and robustness in resource-constrained scenarios.