pith. sign in

arxiv: 2511.18258 · v1 · submitted 2025-11-23 · 💻 cs.MA · cs.AI· cs.LG

Hybrid Agentic AI and Multi-Agent Systems in Smart Manufacturing

Mojtaba A. Farahani , Md Irfan Khan , Thorsten Wuest This is my paper

Pith reviewed 2026-05-17 06:41 UTC · model grok-4.3

classification 💻 cs.MA cs.AIcs.LG
keywords agentic AImulti-agent systemssmart manufacturingprescriptive maintenanceLLM agentshybrid frameworkschema discoveryadaptive intelligence
0
0 comments X

The pith

A hybrid framework lets LLM agents orchestrate schema detection, model adaptation, and prioritized maintenance recommendations while edge agents execute domain tasks in smart manufacturing.

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

The paper establishes that combining agentic AI driven by large language models with traditional multi-agent systems creates a layered architecture capable of handling prescriptive maintenance end-to-end. An LLM Planner Agent manages workflow decisions and context retention across perception, preprocessing, analytics, and optimization layers, while specialized agents autonomously manage schema discovery, feature analysis, model selection, and optimization. Rule-based and small-model agents perform efficient edge tasks, and a human-in-the-loop interface provides transparency and auditability. Sympathetic readers would care because the approach promises dynamic adaptation and interpretable decisions that scale beyond rigid traditional systems, as shown in validation on two industrial datasets where the framework automatically detects schemas, adapts pipelines, optimizes performance, and outputs actionable recommendations. The modular design also supports future extension by adding new agents or modules.

Core claim

The paper presents a hybrid agentic AI and multi-agent framework for prescriptive maintenance in smart manufacturing in which LLM-based agents supply strategic orchestration and adaptive reasoning while rule-based and small language model agents perform efficient, domain-specific tasks on the edge. The framework uses a layered architecture coordinated by an LLM Planner Agent that manages workflow decisions and context retention; specialized agents handle schema discovery, intelligent feature analysis, model selection, and prescriptive optimization; and a HITL interface ensures transparency and auditability of recommendations. Validation on two industrial manufacturing datasets demonstrates a

What carries the argument

The hybrid agentic AI and multi-agent framework with an LLM Planner Agent that coordinates a layered architecture of perception, preprocessing, analytics, and optimization to enable automatic schema detection, pipeline adaptation, and prioritized maintenance recommendations.

If this is right

  • The system automatically detects schemas from industrial datasets without manual setup.
  • Preprocessing pipelines adapt dynamically to the characteristics of incoming data.
  • Model performance improves through agent-driven selection and adaptive intelligence.
  • Actionable and prioritized maintenance recommendations are generated for cost-efficient scheduling.
  • The modular structure supports adding new agents or domain modules as needs evolve.

Where Pith is reading between the lines

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

  • The same planner-plus-specialist pattern could transfer to related manufacturing tasks such as quality inspection or energy optimization by swapping the optimization layer.
  • Tighter coupling of the perception layer to live sensor streams might allow continuous online adaptation beyond the current batch-oriented proof of concept.
  • Scaling the framework across multiple connected factories would expose coordination limits not visible in the two-dataset test.
  • Stronger built-in explainability from the agents could eventually reduce the frequency of HITL reviews in lower-noise settings.

Load-bearing premise

LLM-based agents can reliably perform strategic orchestration and context retention in noisy industrial environments without introducing hallucinations or unsafe recommendations that the HITL interface cannot catch.

What would settle it

Running the framework on a new noisy industrial dataset and checking whether it produces hallucinated or unsafe maintenance recommendations that the HITL interface fails to catch would settle whether the orchestration claim holds.

read the original abstract

The convergence of Agentic AI and MAS enables a new paradigm for intelligent decision making in SMS. Traditional MAS architectures emphasize distributed coordination and specialized autonomy, while recent advances in agentic AI driven by LLMs introduce higher order reasoning, planning, and tool orchestration capabilities. This paper presents a hybrid agentic AI and multi agent framework for a Prescriptive Maintenance use case, where LLM based agents provide strategic orchestration and adaptive reasoning, complemented by rule based and SLMs agents performing efficient, domain specific tasks on the edge. The proposed framework adopts a layered architecture that consists of perception, preprocessing, analytics, and optimization layers, coordinated through an LLM Planner Agent that manages workflow decisions and context retention. Specialized agents autonomously handle schema discovery, intelligent feature analysis, model selection, and prescriptive optimization, while a HITL interface ensures transparency and auditability of generated maintenance recommendations. This hybrid design supports dynamic model adaptation, cost efficient maintenance scheduling, and interpretable decision making. An initial proof of concept implementation is validated on two industrial manufacturing datasets. The developed framework is modular and extensible, supporting seamless integration of new agents or domain modules as capabilities evolve. The results demonstrate the system capability to automatically detect schema, adapt preprocessing pipelines, optimize model performance through adaptive intelligence, and generate actionable, prioritized maintenance recommendations. The framework shows promise in achieving improved robustness, scalability, and explainability for RxM in smart manufacturing, bridging the gap between high level agentic reasoning and low level autonomous execution.

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

2 major / 2 minor

Summary. The manuscript proposes a hybrid agentic AI and multi-agent systems framework for prescriptive maintenance in smart manufacturing. It describes a layered architecture (perception, preprocessing, analytics, optimization) coordinated by an LLM Planner Agent for strategic orchestration and context retention, with specialized agents handling schema discovery, feature analysis, model selection, and prescriptive optimization, plus a human-in-the-loop (HITL) interface for transparency. A proof-of-concept implementation is validated on two industrial manufacturing datasets, with claims that the system can automatically detect schema, adapt preprocessing pipelines, optimize model performance, and generate actionable prioritized maintenance recommendations, promising improved robustness, scalability, and explainability.

Significance. If the empirical claims are substantiated with quantitative evidence, this work could offer a practical bridge between high-level LLM-driven reasoning and low-level autonomous execution in industrial MAS, with the modular design and HITL layer providing a useful template for extensible smart manufacturing systems. The conceptual integration of agentic capabilities with rule-based and SLM agents is a reasonable direction for the field.

major comments (2)
  1. [Validation / Results] Validation / Results: The abstract and PoC description claim that the framework demonstrates automatic schema detection, preprocessing adaptation, model optimization through adaptive intelligence, and generation of prioritized recommendations on two datasets. However, no quantitative metrics (e.g., accuracy, F1 scores, error rates, hallucination frequency), baseline comparisons, success criteria, or details on measurement under noisy sensor conditions are provided. This absence directly undermines verification of the central claims regarding robustness and explainability.
  2. [Framework Architecture] Framework Architecture: The description of the LLM Planner Agent managing workflow decisions and context retention (in the layered architecture) is presented at a high level without specifying fallback mechanisms, context-drift handling, or safeguards against unsafe recommendations. This is load-bearing for the claim that LLM-based strategic orchestration remains reliable in industrial environments.
minor comments (2)
  1. [Abstract] Abstract: The two datasets are referenced but not characterized (e.g., sensor types, noise levels, or domain specifics), which would improve clarity and reproducibility.
  2. [Overall] Notation and Terminology: Terms such as 'adaptive intelligence' and 'prescriptive optimization' are used without precise definitions or distinctions from standard MAS concepts, potentially reducing precision.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive and detailed feedback. We address each major comment below and have revised the manuscript to strengthen the presentation of results and architectural details.

read point-by-point responses

  1. Referee: [Validation / Results] Validation / Results: The abstract and PoC description claim that the framework demonstrates automatic schema detection, preprocessing adaptation, model optimization through adaptive intelligence, and generation of prioritized recommendations on two datasets. However, no quantitative metrics (e.g., accuracy, F1 scores, error rates, hallucination frequency), baseline comparisons, success criteria, or details on measurement under noisy sensor conditions are provided. This absence directly undermines verification of the central claims regarding robustness and explainability.

    Authors: We agree that the initial manuscript presented the proof-of-concept validation primarily through descriptive results without sufficient quantitative metrics or comparisons. In the revised version, we have added a new Evaluation section that reports accuracy, F1 scores, and error rates for the schema detection, feature analysis, and model optimization agents; includes baseline comparisons against non-agentic and rule-based systems; defines explicit success criteria; and presents results from controlled experiments under noisy sensor conditions. These additions directly support the claims of robustness and explainability. revision: yes

  2. Referee: [Framework Architecture] Framework Architecture: The description of the LLM Planner Agent managing workflow decisions and context retention (in the layered architecture) is presented at a high level without specifying fallback mechanisms, context-drift handling, or safeguards against unsafe recommendations. This is load-bearing for the claim that LLM-based strategic orchestration remains reliable in industrial environments.

    Authors: We acknowledge that the original description of the LLM Planner Agent was high-level. The revised manuscript now includes explicit specifications for fallback mechanisms (e.g., automatic delegation to rule-based or SLM agents for safety-critical actions), context-drift handling via periodic consistency checks against incoming sensor data, and safeguards consisting of constraint-based validation of recommendations prior to execution together with mandatory HITL review for high-impact decisions. These details address the reliability requirements for industrial settings. revision: yes

Circularity Check

0 steps flagged

No significant circularity; architectural framework with external PoC validation

full rationale

The manuscript describes a layered hybrid agentic AI/MAS framework for prescriptive maintenance, consisting of perception, preprocessing, analytics, and optimization layers coordinated by an LLM Planner Agent, with specialized agents for schema discovery and model tasks plus a HITL interface. Validation occurs via proof-of-concept implementation on two external industrial manufacturing datasets, demonstrating automatic schema detection, pipeline adaptation, model optimization, and prioritized recommendations. No equations, derivations, fitted parameters, or self-referential definitions appear in the text. Claims rest on the described architecture and empirical PoC results rather than any reduction of outputs to inputs by construction or load-bearing self-citation chains. The framework is self-contained as a modular systems proposal.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

This is an applied systems paper rather than a theoretical derivation. No explicit free parameters, mathematical axioms, or newly postulated physical entities are introduced in the abstract. The design implicitly assumes that current LLM reasoning capabilities are sufficiently reliable for orchestration and that domain-specific rule-based agents can be cleanly separated from high-level planning.

axioms (2)
  • domain assumption LLM agents can perform reliable strategic orchestration and context retention in industrial settings
    Invoked when the LLM Planner Agent is described as managing workflow decisions and context retention.
  • domain assumption Specialized agents can autonomously handle schema discovery, feature analysis, model selection, and optimization without coordination failures
    Stated in the description of autonomous agent responsibilities within the layered architecture.

pith-pipeline@v0.9.0 · 5569 in / 1517 out tokens · 55996 ms · 2026-05-17T06:41:55.811762+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

47 extracted references · 47 canonical work pages · 2 internal anchors

  1. [1]

    Y. Liao, F. Deschamps, E.D.F.R. Loures, L.F.P. Ramos, Past, present and future of Industry 4.0 - a systematic literature review and research agenda proposal, Int. J. Prod. Res. 55 (2017) 3609–3629. https://doi.org/10.1080/00207543.2017.1308576

    work page doi:10.1080/00207543.2017.1308576 2017

  2. [2]

    Farahani, F.E

    M.A. Farahani, F.E. Kalach, A. Harper, M.R. McCormick, R. Harik, T. Wuest, Time-series forecasting in smart manufacturing systems: An experimental evaluation of the state-of-the-art algorithms, Robot. Comput.-Integr. Manuf. 95 (2025) 103010. https://doi.org/10.1016/j.rcim.2025.103010

    work page doi:10.1016/j.rcim.2025.103010 2025

  3. [3]

    O’Donovan, K

    P. O’Donovan, K. Leahy, K. Bruton, D.T.J. O’Sullivan, An industrial big data pipeline for data-driven analytics maintenance applications in large-scale smart manufacturing facilities, J. Big Data 2 (2015) 25. https://doi.org/10.1186/s40537-015-0034-z

    work page doi:10.1186/s40537-015-0034-z 2015

  4. [4]

    J. Wang, C. Xu, J. Zhang, R. Zhong, Big data analytics for intelligent manufacturing systems: A review, J. Manuf. Syst. 62 (2022) 738–752. https://doi.org/10.1016/j.jmsy.2021.03.005

    work page doi:10.1016/j.jmsy.2021.03.005 2022

  5. [5]

    Egyed, P

    A. Egyed, P. Grünbacher, L. Linsbauer, H. Prähofer, I. Schaefer, Variability in Products and Production, in: B. Vogel-Heuser, M. Wimmer (Eds.), Digit. Transform., Springer Berlin Heidelberg, Berlin, Heidelberg, 2023: pp. 65–91. https://doi.org/10.1007/978-3-662-65004- 2_3

    work page doi:10.1007/978-3-662-65004- 2023

  6. [6]

    L. Li, B. Lei, C. Mao, Digital twin in smart manufacturing, J. Ind. Inf. Integr. 26 (2022) 100289. https://doi.org/10.1016/j.jii.2021.100289

    work page doi:10.1016/j.jii.2021.100289 2022

  7. [8]

    Mahale, S

    Y. Mahale, S. Kolhar, A.S. More, A comprehensive review on artificial intelligence driven predictive maintenance in vehicles: technologies, challenges and future research directions, Discov. Appl. Sci. 7 (2025)

    work page 2025

  8. [9]

    https://doi.org/10.1007/s42452-025-06681-3

    work page doi:10.1007/s42452-025-06681-3

  9. [10]

    Ansari, R

    F. Ansari, R. Glawar, T. Nemeth, PriMa: a prescriptive maintenance model for cyber-physical production systems, Int. J. Comput. Integr. Manuf. 32 (2019) 482–503. https://doi.org/10.1080/0951192X.2019.1571236

    work page doi:10.1080/0951192x.2019.1571236 2019

  10. [11]

    Balaji, D

    P.G. Balaji, D. Srinivasan, An Introduction to Multi-Agent Systems, in: D. Srinivasan, L.C. Jain (Eds.), Innov. Multi-Agent Syst. Appl. - 1, Springer Berlin Heidelberg, Berlin, Heidelberg, 2010: pp. 1–27. https://doi.org/10.1007/978-3-642-14435-6_1

    work page doi:10.1007/978-3-642-14435-6_1 2010

  11. [12]

    X. Jing, X. Yao, M. Liu, J. Zhou, Multi-agent reinforcement learning based on graph convolutional network for flexible job shop scheduling, J. Intell. Manuf. 35 (2024) 75–93. https://doi.org/10.1007/s10845-022- 02037-5

    work page doi:10.1007/s10845-022- 2024

  12. [13]

    Kaven, P

    L. Kaven, P. Huke, A. Göppert, R.H. Schmitt, Multi agent reinforcement learning for online layout planning and scheduling in flexible assembly systems, J. Intell. Manuf. 35 (2024) 3917–3936. https://doi.org/10.1007/s10845-023-02309-8

    work page doi:10.1007/s10845-023-02309-8 2024

  13. [15]

    H. Du, S. Thudumu, R. Vasa, K. Mouzakis, A Survey on Context- Aware Multi-Agent Systems: Techniques, Challenges and Future Directions, (2025). https://doi.org/10.48550/arXiv.2402.01968

    work page doi:10.48550/arxiv.2402.01968 2025

  14. [16]

    J. Fang, Y. Peng, X. Zhang, Y. Wang, X. Yi, G. Zhang, Y. Xu, B. Wu, S. Liu, Z. Li, Z. Ren, N. Aletras, X. Wang, H. Zhou, Z. Meng, A Comprehensive Survey of Self-Evolving AI Agents: A New Paradigm Bridging Foundation Models and Lifelong Agentic Systems, (2025). https://doi.org/10.48550/arXiv.2508.07407

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2508.07407 2025

  15. [17]

    Bandi, B

    A. Bandi, B. Kongari, R. Naguru, S. Pasnoor, S.V. Vilipala, The Rise of Agentic AI: A Review of Definitions, Frameworks, Architectures, Applications, Evaluation Metrics, and Challenges, Future Internet 17 (2025) 404. https://doi.org/10.3390/fi17090404

    work page doi:10.3390/fi17090404 2025

  16. [18]

    Roumeliotis, and Manoj Karkee

    R. Sapkota, K.I. Roumeliotis, M. Karkee, AI Agents vs. Agentic AI: A Conceptual taxonomy, applications and challenges, Inf. Fusion 126 (2026) 103599. https://doi.org/10.1016/j.inffus.2025.103599

    work page doi:10.1016/j.inffus.2025.103599 2026

  17. [19]

    B. Yan, K. Li, M. Xu, Y. Dong, Y. Zhang, Z. Ren, X. Cheng, On protecting the data privacy of Large Language Models (LLMs) and LLM agents: A literature review, High-Confid. Comput. 5 (2025) 100300. https://doi.org/10.1016/j.hcc.2025.100300

    work page doi:10.1016/j.hcc.2025.100300 2025

  18. [20]

    Small Language Models are the Future of Agentic AI

    P. Belcak, G. Heinrich, S. Diao, Y. Fu, X. Dong, S. Muralidharan, Y.C. Lin, P. Molchanov, Small Language Models are the Future of Agentic AI, (2025). https://doi.org/10.48550/arXiv.2506.02153

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2506.02153 2025

  19. [21]

    Pulikottil, L.A

    T. Pulikottil, L.A. Estrada-Jimenez, H.U. Rehman, J. Barata, S. Nikghadam-Hojjati, L. Zarzycki, Multi-agent based manufacturing: current trends and challenges, in: 2021 26th IEEE Int. Conf. Emerg. Technol. Fact. Autom. ETFA, IEEE, Vasteras, Sweden, 2021: pp. 1–7. https://doi.org/10.1109/ETFA45728.2021.9613555

    work page doi:10.1109/etfa45728.2021.9613555 2021

  20. [22]

    Fernández-Miguel, S

    A. Fernández-Miguel, S. Ortíz-Marcos, M. Jiménez-Calzado, A.P. Fernández Del Hoyo, F.E. García-Muiña, D. Settembre-Blundo, Agentic AI in Smart Manufacturing: Enabling Human-Centric Predictive Maintenance Ecosystems, Appl. Sci. 15 (2025) 11414. https://doi.org/10.3390/app152111414

    work page doi:10.3390/app152111414 2025

  21. [23]

    Carvalho, F.R.D

    J.N.D. Carvalho, F.R.D. Silva, E.G.S. Nascimento, Challenges of the Biopharmaceutical Industry in the Application of Prescriptive Maintenance in the Industry 4.0 Context: A Comprehensive Literature Review, Sensors 24 (2024) 7163. https://doi.org/10.3390/s24227163

    work page doi:10.3390/s24227163 2024

  22. [24]

    Benhanifia, Z.B

    A. Benhanifia, Z.B. Cheikh, P.M. Oliveira, A. Valente, J. Lima, Systematic review of predictive maintenance practices in the manufacturing sector, Intell. Syst. Appl. 26 (2025) 200501. https://doi.org/10.1016/j.iswa.2025.200501

    work page doi:10.1016/j.iswa.2025.200501 2025

  23. [25]

    M. Pech, J. Vrchota, J. Bednář, Predictive Maintenance and Intelligent Sensors in Smart Factory: Review, Sensors 21 (2021) 1470. https://doi.org/10.3390/s21041470

    work page doi:10.3390/s21041470 2021

  24. [26]

    Theissler, J

    A. Theissler, J. Pérez-Velázquez, M. Kettelgerdes, G. Elger, Predictive maintenance enabled by machine learning: Use cases and challenges in the automotive industry, Reliab. Eng. Syst. Saf. 215 (2021) 107864. https://doi.org/10.1016/j.ress.2021.107864

    work page doi:10.1016/j.ress.2021.107864 2021

  25. [27]

    Lin, Y.-H

    K.-Y. Lin, Y.-H. Hong, M.-H. Li, Y. Shi, K. Matsuno, Predictive maintenance in industrial systems: an XGBoost-based approach for failure time estimation and resource optimization, J. Ind. Prod. Eng. (2025) 1–24. https://doi.org/10.1080/21681015.2025.2519369

    work page doi:10.1080/21681015.2025.2519369 2025

  26. [28]

    Farahani, M.R

    M.A. Farahani, M.R. McCormick, R. Harik, T. Wuest, Time-series classification in smart manufacturing systems: An experimental evaluation of state-of-the-art machine learning algorithms, Robot. Comput.-Integr. Manuf. 91 (2025) 102839. https://doi.org/10.1016/j.rcim.2024.102839

    work page doi:10.1016/j.rcim.2024.102839 2025

  27. [29]

    Hurtado, D

    J. Hurtado, D. Salvati, R. Semola, M. Bosio, V. Lomonaco, Continual learning for predictive maintenance: Overview and challenges, Intell. Syst. Appl. 19 (2023) 200251. https://doi.org/10.1016/j.iswa.2023.200251

    work page doi:10.1016/j.iswa.2023.200251 2023

  28. [30]

    M.I. Khan, M. Rayhan, M.A. Farahani, T. Wuest, Federated Learning for Smart Manufacturing: Evaluating Deep Learning Architectures for Time Series Forecasting in a Collaborative Framework, in: Adv. Prod. Manag. Syst. Cyber-Phys.-Hum. Prod. Syst. Hum.-AI Collab. Beyond, Springer Nature Switzerland, Cham, 2026: pp. 49–61. 12 Farahani et al. Manufacturing Let...

    work page doi:10.1007/978-3-032-03534-9_4 2026

  29. [31]

    Orošnjak, F

    M. Orošnjak, F. Saretzky, S. Kedziora, Prescriptive Maintenance: A Systematic Literature Review and Exploratory Meta-Synthesis, Appl. Sci. 15 (2025) 8507. https://doi.org/10.3390/app15158507

    work page doi:10.3390/app15158507 2025

  30. [32]

    Leitão, Agent-based distributed manufacturing control: A state-of- the-art survey, Eng

    P. Leitão, Agent-based distributed manufacturing control: A state-of- the-art survey, Eng. Appl. Artif. Intell. 22 (2009) 979–991. https://doi.org/10.1016/j.engappai.2008.09.005

    work page doi:10.1016/j.engappai.2008.09.005 2009

  31. [33]

    Leitao, S

    P. Leitao, S. Karnouskos, L. Ribeiro, J. Lee, T. Strasser, A.W. Colombo, Smart Agents in Industrial Cyber–Physical Systems, Proc. IEEE 104 (2016) 1086–1101. https://doi.org/10.1109/JPROC.2016.2521931

    work page doi:10.1109/jproc.2016.2521931 2016

  32. [34]

    Jennings, K

    N.R. Jennings, K. Sycara, M. Wooldridge, A Roadmap of Agent Research and Development, Auton. Agents Multi-Agent Syst. 1 (1998) 7–38. https://doi.org/10.1023/A:1010090405266

    work page doi:10.1023/a:1010090405266 1998

  33. [35]

    Pal, M.L

    M. Pal, M.L. Mittal, G. Soni, S.S. Chouhan, A multi-agent system for integrated scheduling and maintenance planning of the flexible job shop, Comput. Oper. Res. 159 (2023) 106365. https://doi.org/10.1016/j.cor.2023.106365

    work page doi:10.1016/j.cor.2023.106365 2023

  34. [36]

    J. Lim, L. Pfeiffer, F. Ocker, B. Vogel-Heuser, I. Kovalenko, Ontology- Based Feedback to Improve Runtime Control for Multi-Agent Manufacturing Systems, (2023). https://doi.org/10.48550/arXiv.2309.10132

    work page doi:10.48550/arxiv.2309.10132 2023

  35. [37]

    C. Li, Q. Chang, H.-T. Fan, Multi-agent reinforcement learning for integrated manufacturing system-process control, J. Manuf. Syst. 76 (2024) 585–598. https://doi.org/10.1016/j.jmsy.2024.08.021

    work page doi:10.1016/j.jmsy.2024.08.021 2024

  36. [38]

    J. Lim, B. Vogel-Heuser, I. Kovalenko, Large Language Model- Enabled Multi-Agent Manufacturing Systems, (2024). https://doi.org/10.48550/arXiv.2406.01893

    work page doi:10.48550/arxiv.2406.01893 2024

  37. [39]

    J. Lim, I. Kovalenko, A Large Language Model-Enabled Control Architecture for Dynamic Resource Capability Exploration in Multi- Agent Manufacturing Systems, (2025). https://doi.org/10.48550/arXiv.2505.22814

    work page doi:10.48550/arxiv.2505.22814 2025

  38. [40]

    Joshi, BlockPay: A Blockchain-Based Framework for Secure and Scalable Digital Payments, Int

    S. Joshi, BlockPay: A Blockchain-Based Framework for Secure and Scalable Digital Payments, Int. J. Innov. Res. Eng. Manag. 12 (2025) 65–76. https://doi.org/10.55524/ijirem.2025.12.3.9

    work page doi:10.55524/ijirem.2025.12.3.9 2025

  39. [41]

    Jiang, F

    W. Jiang, F. Hu, Artificial Intelligence Agent-Enabled Predictive Maintenance: Conceptual Proposal and Basic Framework, Computers 14 (2025) 329. https://doi.org/10.3390/computers14080329

    work page doi:10.3390/computers14080329 2025

  40. [42]

    S. Raza, R. Sapkota, M. Karkee, C. Emmanouilidis, TRiSM for Agentic AI: A Review of Trust, Risk, and Security Management in LLM-based Agentic Multi-Agent Systems, (2025). https://doi.org/10.48550/arXiv.2506.04133

    work page doi:10.48550/arxiv.2506.04133 2025

  41. [43]

    Z. Duan, J. Wang, Exploration of LLM Multi-Agent Application Implementation Based on LangGraph+CrewAI, (2024). https://doi.org/10.48550/arXiv.2411.18241

    work page doi:10.48550/arxiv.2411.18241 2024

  42. [44]

    Agentic AI Frameworks: Architectures, Protocols, and Design Challenges, August 2025

    H. Derouiche, Z. Brahmi, H. Mazeni, Agentic AI Frameworks: Architectures, Protocols, and Design Challenges, (2025). https://doi.org/10.48550/ARXIV.2508.10146

    work page doi:10.48550/arxiv.2508.10146 2025

  43. [45]

    L. Sun, S. Fu, B. Yao, Y. Lu, W. Li, H. Gu, J. Gesi, J. Huang, C. Luo, D. Wang, LLM Agent Meets Agentic AI: Can LLM Agents Simulate Customers to Evaluate Agentic-AI-based Shopping Assistants?, (2025). https://doi.org/10.48550/arXiv.2509.21501

    work page doi:10.48550/arxiv.2509.21501 2025

  44. [46]

    L. Wang, C. Ma, X. Feng, Z. Zhang, H. Yang, J. Zhang, Z. Chen, J. Tang, X. Chen, Y. Lin, W.X. Zhao, Z. Wei, J.-R. Wen, A Survey on Large Language Model based Autonomous Agents, Front. Comput. Sci. 18 (2024) 186345. https://doi.org/10.1007/s11704-024-40231-1

    work page doi:10.1007/s11704-024-40231-1 2024

  45. [47]

    Dimeas, N.D

    A.L. Dimeas, N.D. Hatziargyriou, Multi-agent reinforcement learning for microgrids, in: IEEE PES Gen. Meet., IEEE, Minneapolis, MN, 2010: pp. 1–8. https://doi.org/10.1109/PES.2010.5589633

    work page doi:10.1109/pes.2010.5589633 2010

  46. [48]

    Belhadi, Y

    A. Belhadi, Y. Djenouri, G. Srivastava, J.C.-W. Lin, Reinforcement learning multi-agent system for faults diagnosis of mircoservices in industrial settings, Comput. Commun. 177 (2021) 213–219. https://doi.org/10.1016/j.comcom.2021.07.010

    work page doi:10.1016/j.comcom.2021.07.010 2021

  47. [49]

    W. Luo, T. Hu, Y. Ye, C. Zhang, Y. Wei, A hybrid predictive maintenance approach for CNC machine tool driven by Digital Twin, Robot. Comput.-Integr. Manuf. 65 (2020) 101974. https://doi.org/10.1016/j.rcim.2020.101974

    work page doi:10.1016/j.rcim.2020.101974 2020