pith. machine review for the scientific record. sign in

arxiv: 2605.08831 · v1 · submitted 2026-05-09 · 💻 cs.RO

Recognition: no theorem link

AssemPlanner: A Multi-Agent Based Task Planning Framework for Flexible Assembly System

Chaoran Zhang, Chenhao Zhang, Long Zeng, Pingfa Feng, Yongbo Yang, Zhaobo Xu

Pith reviewed 2026-05-12 01:32 UTC · model grok-4.3

classification 💻 cs.RO
keywords multi-agent systemstask planningflexible assemblyReActscene graphindustrial roboticsnatural language input
0
0 comments X

The pith

AssemPlanner uses a ReAct-based multi-agent system to convert natural language task descriptions into sequential assembly operations by adapting to feedback from specialized agents and a scene graph.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper aims to establish that a multi-agent framework can automate task planning in flexible assembly systems from natural language inputs alone. Existing approaches demand lengthy expert configuration to set up a production line for each new product, so the goal is to bypass that bottleneck. The framework relies on a central SchedAgent that follows the ReAct pattern to revise its decisions dynamically after receiving input from KnowledgeAgent, LineBalanceAgent, and the scene graph. This loop is intended to resolve the detailed constraints of industrial processes without further human input. If the approach holds, production lines could be reconfigured for different products far more rapidly than before.

Core claim

AssemPlanner is a multi-agent based task planning framework for flexible assembly systems that accepts tasks described in natural language and transforms them into actionable sequential production operations. It includes specialized agents such as SchedAgent, KnowledgeAgent, LineBalanceAgent, and a scene graph, with the SchedAgent serving as the central ReAct-based reasoning engine that adaptively adjusts actions through multi-agent feedback to autonomously resolve complex industrial process constraints.

What carries the argument

The ReAct-based SchedAgent that observes feedback from KnowledgeAgent, LineBalanceAgent, and the scene graph to adaptively adjust its planning actions and resolve constraints.

If this is right

  • Natural language task descriptions can be converted directly into production plans without multi-expert setup time.
  • Complex industrial constraints can be resolved autonomously through iterative multi-agent feedback.
  • Flexible assembly systems can switch between products with reduced reconfiguration effort.
  • Public release of code and datasets enables direct testing and extension by others.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same feedback-driven structure could transfer to task planning problems in other robotic domains such as disassembly or maintenance.
  • Incorporating additional domain-specific agents might extend the method to larger or less structured factory settings.
  • Deployment on physical hardware would reveal how the scene graph and feedback loop handle real-world sensor noise or timing variations.

Load-bearing premise

The multi-agent feedback loop from the scene graph and specialized agents can accurately capture and resolve all relevant industrial process constraints without additional human configuration or intervention.

What would settle it

Supply a previously unseen assembly task in natural language to AssemPlanner and check whether it outputs a complete, feasible production sequence that satisfies all constraints with zero manual edits or expert additions.

read the original abstract

In flexible assembly systems, existing task planning methods require a time-consuming configuration process by multiple experts to establish a production line for a new product. To address this challenge, we propose a multi-agent based task planning framework for flexible assembly systems, denoted as AssemPlanner. It takes tasks described in natural language as input, which are then converted into actionable sequential production operations. It comprises several specialized agents, including SchedAgent , KnowledgeAgent, LineBalanceAgent, and a scene graph. Within the proposed framework, SchedAgent serves as the central reasoning engine. Departing from traditional static pipelines, AssemPlanner utilizes a ReAct-based SchedAgent to adaptively adjust actions via multi-agent feedback. By observing the feedback from KnowledgeAgent, LineBalanceAgent, and the scene graph, it autonomously resolves complex industrial process constraints. To facilitate reproducibility, all code and datasets are released at https://github.com/chz332/Assemplanner.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 1 minor

Summary. The manuscript proposes AssemPlanner, a multi-agent task planning framework for flexible assembly systems. Natural-language task inputs are converted to sequential production operations via specialized agents (SchedAgent as the ReAct-based central engine, KnowledgeAgent, LineBalanceAgent, and a scene graph). The core claim is that multi-agent feedback enables the SchedAgent to adaptively adjust actions and autonomously resolve industrial constraints (precedence, resource allocation, line balancing, safety) without the time-consuming expert configuration required by prior static pipelines. Code and datasets are released for reproducibility.

Significance. If the autonomous constraint-resolution claim holds, the work could meaningfully reduce setup time for new products in flexible manufacturing by replacing expert-driven configuration with LLM-based multi-agent reasoning. The open release of code and datasets is a clear strength that supports reproducibility and future extensions. However, the current lack of any empirical validation means the assessed significance is prospective rather than demonstrated.

major comments (3)
  1. [Abstract] Abstract: the central claim that the ReAct-based SchedAgent 'autonomously resolves complex industrial process constraints' by observing feedback from KnowledgeAgent, LineBalanceAgent, and the scene graph is load-bearing for the paper's contribution yet is unsupported by any experiments, case studies, or quantitative results showing successful resolution of precedence, resource, balancing, or safety constraints.
  2. [Section 3] Section 3 (Framework): no formal enumeration or specification of the constraint language is provided, nor any argument or proof that the agents plus scene graph achieve complete coverage of all relevant industrial constraints; without this, the completeness assumption underlying autonomous resolution cannot be assessed.
  3. [Section 4] Section 4 (Evaluation, if present): the manuscript contains no ablation studies, failure-mode analysis, or comparisons against baselines that would demonstrate whether the ReAct feedback loop reliably detects and repairs violations or requires external intervention when a constraint is omitted from the feedback channels.
minor comments (1)
  1. [Abstract] Abstract and Section 3: the high-level description of agent interactions could be clarified with a diagram or pseudocode showing the exact ReAct loop, observation format, and action space.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback on our manuscript. We address each major comment point by point below, clarifying the scope of our contribution as a framework proposal while agreeing that additional elements would strengthen the presentation. We plan revisions to incorporate clarifications and illustrative examples.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the ReAct-based SchedAgent 'autonomously resolves complex industrial process constraints' by observing feedback from KnowledgeAgent, LineBalanceAgent, and the scene graph is load-bearing for the paper's contribution yet is unsupported by any experiments, case studies, or quantitative results showing successful resolution of precedence, resource, balancing, or safety constraints.

    Authors: We acknowledge that the manuscript does not contain experiments, case studies, or quantitative results to empirically validate the autonomous resolution of constraints. The central claim describes the intended operation of the ReAct-based SchedAgent, which adaptively adjusts plans by incorporating feedback from the specialized agents and scene graph as detailed in the framework design. This is presented as an architectural approach to reduce expert configuration rather than a validated system. In the revised manuscript, we will add qualitative illustrative scenarios demonstrating constraint resolution through the feedback loop. revision: yes

  2. Referee: [Section 3] Section 3 (Framework): no formal enumeration or specification of the constraint language is provided, nor any argument or proof that the agents plus scene graph achieve complete coverage of all relevant industrial constraints; without this, the completeness assumption underlying autonomous resolution cannot be assessed.

    Authors: We agree that an explicit enumeration of addressed constraints would improve clarity. The framework targets practical constraints in flexible assembly: precedence and resource knowledge via KnowledgeAgent, line balancing via LineBalanceAgent, and spatial/safety via scene graph feedback to the SchedAgent. We do not claim or prove exhaustive coverage of every possible industrial constraint, as the system is scoped to common assembly scenarios. In revision, we will add a dedicated subsection enumerating the constraint types handled by each agent and discussing the intended scope. revision: yes

  3. Referee: [Section 4] Section 4 (Evaluation, if present): the manuscript contains no ablation studies, failure-mode analysis, or comparisons against baselines that would demonstrate whether the ReAct feedback loop reliably detects and repairs violations or requires external intervention when a constraint is omitted from the feedback channels.

    Authors: The manuscript is structured as a framework proposal and therefore does not include an evaluation section with ablations, failure-mode analysis, or baseline comparisons. The focus is on the novel integration of ReAct reasoning with multi-agent feedback for adaptive planning. We recognize that such analyses would help assess the reliability of the feedback mechanism. We will add a new section with example-driven discussion of the feedback loop, including potential failure modes and when external intervention might be needed. revision: yes

Circularity Check

0 steps flagged

No circularity: architectural proposal without derivations or self-referential reductions

full rationale

The paper describes a proposed multi-agent system (AssemPlanner) using ReAct-based SchedAgent with feedback from KnowledgeAgent, LineBalanceAgent, and scene graph. No equations, derivations, predictions, or fitted parameters appear in the provided text. Central claims are statements of intended system behavior rather than results obtained by reducing inputs to outputs via construction. No self-citations are invoked as load-bearing uniqueness theorems or ansatzes. The architecture is self-contained as a design proposal; any completeness assumptions about constraint resolution are empirical claims, not circular reductions.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 4 invented entities

The central claim depends on the effectiveness of these agents and the scene graph in providing reliable feedback, which is assumed rather than demonstrated in the abstract.

axioms (1)
  • domain assumption Specialized agents can provide accurate and useful feedback on assembly constraints and line balancing.
    This is necessary for the SchedAgent to adaptively resolve issues as described.
invented entities (4)
  • SchedAgent no independent evidence
    purpose: Central ReAct-based reasoning engine for task planning
    New component introduced in the framework.
  • KnowledgeAgent no independent evidence
    purpose: Provides task knowledge
    Specialized agent in the system.
  • LineBalanceAgent no independent evidence
    purpose: Handles line balancing
    Specialized agent in the system.
  • scene graph no independent evidence
    purpose: Represents the assembly environment
    Component of the framework.

pith-pipeline@v0.9.0 · 5473 in / 1673 out tokens · 73421 ms · 2026-05-12T01:32:06.872954+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references · 50 canonical work pages · 8 internal anchors

  1. [1]

    Z. Xu, C. Zhang, S. Hu, Z. Han, P. Feng, L. Zeng, Reconfigurable flexible assembly model and implementation for cross-category products, Journal of Manufacturing Systems, 77 (2024) 154-169

    work page 2024

  2. [2]

    S. Hong, M. Zhuge, J. Chen, X. Zheng, Y. Cheng, C. Zhang, J. Wang, Z. Wang, S.K.S. Yau, Z. Lin, L. Zhou, C. Ran, L. Xiao, C. Wu, J. Schmidhuber, MetaGPT: Meta Programming for A Multi-Agent Collaborative Framework, in, 2023, pp. arXiv:2308.00352

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  3. [3]

    Q. Wu, G. Bansal, J. Zhang, Y. Wu, B. Li, E. Zhu, L. Jiang, X. Zhang, S. Zhang, J. Liu, A.H. Awadallah, R.W. White, D. Burger, C. Wang, AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation, in, 2023, pp. arXiv:2308.08155

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  4. [4]

    Leo Kumar, Knowledge-based expert system in manufacturing planning: state-of-the-art review, International Journal of Production Research, 57 (2019) 4766-4790

    S.P. Leo Kumar, Knowledge-based expert system in manufacturing planning: state-of-the-art review, International Journal of Production Research, 57 (2019) 4766-4790

    work page 2019

  5. [5]

    Z. Liu, Q. Liu, W. Xu, L. Wang, Z. Zhou, Robot learning towards smart robotic manufacturing: A review, Robotics and Computer-Integrated Manufacturing, 77 (2022) 102360

    work page 2022

  6. [6]

    L. Ren, J. Dong, S. Liu, L. Zhang, L. Wang, Embodied Intelligence Toward Future Smart Manufacturing in the Era of AI Foundation Model, IEEE/ASME Transactions on Mechatronics, 30 (2025) 2632-2642

    work page 2025

  7. [7]

    Z. Hu, X. Li, X. Pan, S. Wen, J. Bao, A question answering system for assembly process of wind turbines based on multi-modal knowledge graph and large language model, Journal of Engineering Design, 36 (2025) 1093-1117

    work page 2025

  8. [8]

    Yaoxian, S

    S. Yaoxian, S. Penglei, L. Haoyu, L. Zhixu, S. Wei, X. Yanghua, Z. Xiaofang, Scene-Driven Multimodal Knowledge Graph Construction for Embodied AI, in, 2023, pp. arXiv:2311.03783

    work page arXiv 2023

  9. [9]

    H. Choi, J. Jeong, A Conceptual Framework for a Latest Information-Maintaining Method Using Retrieval-Augmented Generation and a Large Language Model in Smart Manufacturing: Theoretical Approach and Performance Analysis, Machines, 13 (2025) 94. 32

    work page 2025

  10. [10]

    S. Hu, Z. Xu, L. Xiao, P. Feng, L. Zeng, Research on Reconfigurable and Flexible Assembly Technology for Cross-category Products Journal of Mechanical Engineering 60 (2024) 69-81

    work page 2024

  11. [11]

    Zhang, C

    C. Zhang, C. Zhang, Z. Xu, Q. Xie, J. Hou, P. Feng, L. Zeng, Embodied intelligent industrial robotics: Concepts and techniques, in, 2025, pp. arXiv:2505.09305

    work page arXiv 2025

  12. [12]

    L. Xiao, L. Zeng, Z. Xu, X. Liu, Assembly language design and development for reconfigurable flexible assembly line, Robotics and Computer-Integrated Manufacturing, 80 (2023) 102467

    work page 2023

  13. [13]

    Z. Xu, C. Zhang, S. Hou, Z. Han, L. Zeng, P. Feng, Assembly task planning framework based on knowledge graph, Journal of Intelligent Manufacturing, (2025)

    work page 2025

  14. [14]

    Z. Guo, L. Xia, Y. Yu, T. Ao, C. Huang, LightRAG: Simple and Fast Retrieval-Augmented Generation, in, 2024, pp. arXiv:2410.05779

    work page internal anchor Pith review arXiv 2024

  15. [15]

    Y. Gao, Y. Xiong, X. Gao, K. Jia, J. Pan, Y. Bi, Y. Dai, J. Sun, M. Wang, H. Wang, Retrieval-Augmented Generation for Large Language Models: A Survey, in, 2023, pp. arXiv:2312.10997

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  16. [16]

    H. Han, Y. Wang, H. Shomer, K. Guo, J. Ding, Y. Lei, M. Halappanavar, R.A. Rossi, S. Mukherjee, X. Tang, Q. He, Z. Hua, B. Long, T. Zhao, N. Shah, A. Javari, Y. Xia, J. Tang, Retrieval-Augmented Generation with Graphs (GraphRAG), in, 2024, pp. arXiv:2501.00309

    work page arXiv 2024

  17. [17]

    Reflexion: Language Agents with Verbal Reinforcement Learning

    N. Shinn, F. Cassano, E. Berman, A. Gopinath, K. Narasimhan, S. Yao, Reflexion: Language Agents with Verbal Reinforcement Learning, in, 2023, pp. arXiv:2303.11366

    work page internal anchor Pith review Pith/arXiv arXiv 2023

  18. [18]

    S. Yao, J. Zhao, D. Yu, N. Du, I. Shafran, K. Narasimhan, Y. Cao, ReAct: Synergizing Reasoning and Acting in Language Models, in, 2022, pp. arXiv:2210.03629

    work page internal anchor Pith review Pith/arXiv arXiv 2022

  19. [19]

    Buchgeher, D

    G. Buchgeher, D. Gabauer, J. Martinez-Gil, L. Ehrlinger, Knowledge Graphs in Manufacturing and Production: A Systematic Literature Review, IEEE Access, 9 (2021) 55537-55554

    work page 2021

  20. [20]

    Merdan, E

    M. Merdan, E. List, W. Lepuschitz, Knowledge-driven industrial robotics for flexible production, in: 2017 IEEE 21st International Conference on Intelligent Engineering Systems (INES), 2017, pp. 000225-000230

    work page 2017

  21. [21]

    Hoebert, W

    T. Hoebert, W. Lepuschitz, M. Vincze, M. Merdan, Knowledge-driven framework for industrial robotic systems, Journal of Intelligent Manufacturing, 34 (2023) 771-788

    work page 2023

  22. [22]

    Jiang, C

    Y. Jiang, C. Chen, X. Liu, Assembly Process Knowledge Graph for Digital Twin, in: 2021 IEEE 17th International Conference on Automation Science and Engineering (CASE), 2021, pp. 758-763

    work page 2021

  23. [23]

    B. Zhou, J. Bao, Z. Chen, Y. Liu, KGAssembly: Knowledge graph-driven assembly process generation and evaluation for complex components, International Journal of Computer Integrated Manufacturing, 35 (2022) 1151-1171

    work page 2022

  24. [24]

    Z. Qin, Y. Lu, A Knowledge Graph-based knowledge representation for adaptive manufacturing control under mass personalization, Manufacturing Letters, 35 (2023) 96-104. 33

    work page 2023

  25. [25]

    Smirnov, F

    P. Smirnov, F. Joublin, A. Ceravola, M. Gienger, Generating consistent PDDL domains with Large Language Models, in, 2024, pp. arXiv:2404.07751

    work page arXiv 2024

  26. [26]

    Kootbally, C

    Z. Kootbally, C. Schlenoff, C. Lawler, T. Kramer, S.K. Gupta, Towards robust assembly with knowledge representation for the planning domain definition language (PDDL), Robotics and Computer-Integrated Manufacturing, 33 (2015) 42-55

    work page 2015

  27. [27]

    Rogalla, A

    A. Rogalla, A. Fay, O. Niggemann, Improved Domain Modeling for Realistic Automated Planning and Scheduling in Discrete Manufacturing, in: 2018 IEEE 23rd International Conference on Emerging Technologies and Factory Automation (ETFA), 2018, pp. 464-471

    work page 2018

  28. [28]

    C. Li, P. Zheng, Y. Yin, B. Wang, L. Wang, Deep reinforcement learning in smart manufacturing: A review and prospects, CIRP Journal of Manufacturing Science and Technology, 40 (2023) 75-101

    work page 2023

  29. [29]

    Zhang, L.Z

    H. Zhang, L.Z. Liu, H. Xie, Y. Jiang, J. Zhou, Y. Wang, Deep Learning-Based Robot Vision: High-End Tools for Smart Manufacturing, IEEE Instrumentation & Measurement Magazine, 25 (2022) 27-35

    work page 2022

  30. [30]

    F. Fan, G. Xu, N. Feng, L. Li, W. Jiang, L. Yu, X. Xiong, Spatiotemporal path tracking via deep reinforcement learning of robot for manufacturing internal logistics, Journal of Manufacturing Systems, 69 (2023) 150-169

    work page 2023

  31. [31]

    Jiang, Y

    M. Jiang, Y. Guo, S. Huang, J. Pu, L. Zhang, S. Wang, A novel fine-grained assembly sequence planning method based on knowledge graph and deep reinforcement learning, Journal of Manufacturing Systems, 76 (2024) 371-384

    work page 2024

  32. [32]

    Wang, Y.D

    Y.Q. Wang, Y.D. Hu, S.E. Zaatari, W.D. Li, Y. Zhou, Optimised Learning from Demonstrations for Collaborative Robots, Robotics and Computer-Integrated Manufacturing, 71 (2021) 102169

    work page 2021

  33. [33]

    Scherzinger, A

    S. Scherzinger, A. Roennau, R. Dillmann, Contact Skill Imitation Learning for Robot-Independent Assembly Programming, in: 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2019, pp. 4309-4316

    work page 2019

  34. [34]

    Tanaka, S

    Y. Tanaka, S. Katsura, A Voice-Controlled Motion Reproduction Using Large Language Models for Polishing Robots, in: 2023 IEEE International Conference on Mechatronics (ICM), 2023, pp. 1-6

    work page 2023

  35. [35]

    T. Wang, J. Fan, P. Zheng, An LLM-based vision and language cobot navigation approach for Human-centric Smart Manufacturing, Journal of Manufacturing Systems, 75 (2024) 299-305

    work page 2024

  36. [36]

    Fakih, R

    M. Fakih, R. Dharmaji, Y. Moghaddas, G.Q. Araya, O. Ogundare, M.A.A. Faruque, LLM4PlC: Harnessing large Language Models for Verifiable Programming of PlCs in Industrial Control Systems, in: 2024 IEEE/ACM 46th International Conference on Software Engineering: Software Engineering in Practice (ICSE-SEIP), 2024, pp. 192-203

    work page 2024

  37. [37]

    H. Fan, X. Liu, J.Y.H. Fuh, W.F. Lu, B. Li, Embodied intelligence in manufacturing: leveraging large language models for autonomous industrial robotics, Journal of Intelligent Manufacturing, 36 (2025) 1141-1157

    work page 2025

  38. [38]

    Y. Gan, B. Zhang, J. Shao, Z. Han, A. Li, X. Dai, Embodied Intelligence: Bionic Robot Controller Integrating Environment Perception, Autonomous Planning, and Motion Control, IEEE Robotics and Automation Letters, 9 (2024) 4559-4566. 34

    work page 2024

  39. [39]

    S. Han, Q. Zhang, Y. Yao, W. Jin, Z. Xu, LLM Multi-Agent Systems: Challenges and Open Problems, in, 2024, pp. arXiv:2402.03578

    work page arXiv 2024

  40. [40]

    X. Li, S. Wang, S. Zeng, Y. Wu, Y. Yang, A survey on LLM-based multi-agent systems: workflow, infrastructure, and challenges, Vicinagearth, 1 (2024) 9

    work page 2024

  41. [41]

    C. Qian, W. Liu, H. Liu, N. Chen, Y. Dang, J. Li, C. Yang, W. Chen, Y. Su, X. Cong, J. Xu, D. Li, Z. Liu, M. Sun, ChatDev: Communicative Agents for Software Development, in, 2023, pp. arXiv:2307.07924

    work page internal anchor Pith review arXiv 2023

  42. [42]

    HyperAgent: Generalist software engineering agents to solve coding tasks at scale.arXiv:2409.16299, 2024

    H. Nhat Phan, T.N. Nguyen, P.X. Nguyen, N.D.Q. Bui, HyperAgent: Generalist Software Engineering Agents to Solve Coding Tasks at Scale, in, 2024, pp. arXiv:2409.16299

    work page arXiv 2024

  43. [43]

    Magentic-one: A generalist multi- agent system for solving complex tasks.arXiv preprint arXiv:2411.04468, 2024

    A. Fourney, G. Bansal, H. Mozannar, C. Tan, E. Salinas, Erkang, Zhu, F. Niedtner, G. Proebsting, G. Bassman, J. Gerrits, J. Alber, P. Chang, R. Loynd, R. West, V. Dibia, A. Awadallah, E. Kamar, R. Hosn, S. Amershi, Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks, in, 2024, pp. arXiv:2411.04468

    work page arXiv 2024

  44. [44]

    B. Liu, Y. Jiang, X. Zhang, Q. Liu, S. Zhang, J. Biswas, P. Stone, LLM+P: Empowering Large Language Models with Optimal Planning Proficiency, in, 2023, pp. arXiv:2304.11477

    work page internal anchor Pith review arXiv 2023

  45. [45]

    J. Jeon, Y. Sim, H. Lee, C. Han, D. Yun, E. Kim, S.L. Nagendra, M.B.G. Jun, Y. Kim, S.W. Lee, J. Lee, ChatCNC: Conversational machine monitoring via large language model and real-time data retrieval augmented generation, Journal of Manufacturing Systems, 79 (2025) 504-514

    work page 2025

  46. [46]

    Gautam, M.R

    A. Gautam, M.R. Aryal, S. Deshpande, S. Padalkar, M. Nikolaenko, M. Tang, S. Anand, IIoT-enabled digital twin for legacy and smart factory machines with LLM integration, Journal of Manufacturing Systems, 80 (2025) 511-523

    work page 2025

  47. [47]

    Z. Wang, C. Wan, J. Liu, X. Zhang, H. Wang, Y. Hu, Z. Hu, MASC: Large language model-based multi-agent scheduling chain for flexible job shop scheduling problem, Advanced Engineering Informatics, 67 (2025) 103527

    work page 2025

  48. [48]

    Y. Sun, Q. Zhang, J. Bao, Y. Lu, S. Liu, Empowering digital twins with large language models for global temporal feature learning, Journal of Manufacturing Systems, 74 (2024) 83-99

    work page 2024

  49. [49]

    C. Liu, J. Song, D. Tang, L. Wang, H. Zhu, Q. Cai, Probing a novel machine tool fault reasoning and maintenance service recommendation approach through data-knowledge empowered LLMs integrated with AR-assisted maintenance guidance, Advanced Engineering Informatics, 66 (2025) 103460

    work page 2025

  50. [50]

    Z. Ni, X. Deng, C. Tai, X. Zhu, Q. Xie, W. Huang, X. Wu, L. Zeng, GRID: Scene-Graph-based Instruction-driven Robotic Task Planning, in: 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2024, pp. 13765-13772

    work page 2024