Large Language Model Agents in Manufacturing Survey

Language models have given manufacturing agents something they never had: the ability to read, explain, and improvise. This survey organized the resulting literature into four categories of increasing autonomy, compared the architectural ideas that define each, and scrutinized the evidence behind reported capabilities. The pattern is consistent: knowledge access is approaching practical maturity; decision support is useful but poorly measured; autonomous orchestration is architecturally inventive and empirically thin; and interfaces are promising wherever interpretation errors cannot reach actuators. The field's frontier is therefore not greater model capability but engineered trustworthiness—grounding, verification, and evaluation infrastructure that lets an improvising agent be safely wrong. The companion challenge for the research community is to build and study reference architectures with exactly that property.

[1] A. Vaswani et al., “Attention is all you need,” in Advances in Neural Information Processing Systems (NeurIPS), 2017. [2] J. Wei et al., “Chain-of-thought prompting elicits reasoning in large language models,” in Advances in Neural Information Processing Systems (NeurIPS), 2022. [3] S. Yao et al., “ReAct: Synergizing reasoning and acting in language models,” in Proc. International Conference on Learning Representations (ICLR), 2023. [4] P. Lewis et al., “Retrieval-augmented generation for knowledge-intensive NLP tasks,” in Advances in Neural Information Processing Systems (NeurIPS), 2020. [5] P. Leitão, “Agent-based distributed manufacturing control: A state-of-the-art survey,” Engineering Applications of Artificial Intelligence, vol. 22, no. 7, pp. 979–991, 2009. [6] T. Wuest, D. Weimer, C. Irgens, and K.-D. Thoben, “Machine learning in manufacturing: advantages, challenges, and applications,” Production & Manufacturing Research, vol. 4, no. 1, pp. 23–45, 2016. [7] X. Xu, Y. Lu, B. Vogel-Heuser, and L. Wang, “Industry 4.0 and Industry 5.0—Inception, conception and perception,” Journal of Manufacturing Systems, vol. 61, pp. 530–535, 2021. [8] “A survey of emerging applications of large language models for problems in mechanics, product design, and manufacturing,” Advanced Engineering Informatics, 2025. [Online]. Available: https://www.sciencedirect.com/science/article/abs/pii/S1474034624007171 * [9] “Foundation-Model-Based Agents in Industrial Automation: Purposes, Capabilities, and Open Challenges,” arXiv preprint, 2026. [Online]. Available: https://arxiv.org/pdf/2605.02592 * [10] “LLM-Powered AI Agent Systems and Their Applications in Industry,” arXiv preprint, 2025. [Online]. Available: https://arxiv.org/pdf/2505.16120 *