Foundation models and intelligent agents in chemical engineering: advances, challenges, and perspectives

Jibin Zhou , Feiyang Xu , Jinyu Sun , Xin Li , Mao Ye

ENG. Chem. Eng. ›› 2027, Vol. 21 ›› Issue (2) : 16

PDF (1058KB)
ENG. Chem. Eng. ›› 2027, Vol. 21 ›› Issue (2) :16 DOI: 10.1007/s11705-027-2724-4
VIEWS & COMMENTS
Foundation models and intelligent agents in chemical engineering: advances, challenges, and perspectives
Author information +
History +
PDF (1058KB)

Abstract

Chemical engineering spans scales of molecules, catalysts, reactors, units, and plants, governed by physical laws and subject to stringent safety, operability, and reliability requirements. General-purpose large language models, although increasingly capable of scientific reasoning, often lack the domain grounding, physical consistency, and engineering reliability required for industrial deployment. We argue that reliable chemical engineering AI will require the closed-loop integration of domain-specific foundation models, agentic orchestration, and physics- and simulation-based validation. Within this framework, domain-specific models provide knowledge for task-specific reasoning, multi-agent systems support complex task orchestration and specialized tool use, and mechanistic models, engineering constraints, and process simulators assess feasibility and provide feedback for refinement. Existing studies largely demonstrate individual components rather than a fully integrated engineering intelligence lifecycle. This Perspective reviews recent progress in domain-specific foundation models, agent-based process design and optimization, process safety analysis, and operational decision support. We outline challenges and opportunities across three dimensions: strengthening the domain-specific foundation models, developing coordinated agent platforms for multiscale engineering workflows, and establishing traceable evidence and controlled updating mechanisms that distinguish task-level feedback, offline version updates, and online adaptation. Together, these directions define a pathway from conversational chemical engineering models toward reliable, verifiable, and continuously improving engineering intelligence.

Graphical abstract

Keywords

chemical engineering / domain-specific foundation models / multi-agent systems / closed-loop validation / engineering intelligence

Cite this article

Download citation ▾
Jibin Zhou, Feiyang Xu, Jinyu Sun, Xin Li, Mao Ye. Foundation models and intelligent agents in chemical engineering: advances, challenges, and perspectives. ENG. Chem. Eng., 2027, 21 (2) : 16 DOI:10.1007/s11705-027-2724-4

登录浏览全文

4963

注册一个新账户 忘记密码

References

[1]

Li J H , Kwauk M . Exploring complex systems in chemical engineering—the multi-scale methodology. Chemical Engineering Science, 2003, 58(3): 521–535

[2]

Charpentier J C. . The triplet ‘molecular processes-product-process’ engineering: the future of chemical engineering?. Chemical Engineering Science, 2002, 57(22): 4667–4690

[3]

Joly J F , Giroudière F , Bertoncini F . Innovation in process development: from catalyst to industrial process. Catalysis Today, 2013, 218: 153–161

[4]

Jabbar A, Islam S U, Boudjadar J. A comparative review of LLM-based conversational systems: insights from DeepSeek, ChatGPT, Gemini, Claude, and Copilot. In: International Conference on AI and the Digital Economy (CADE 2025). Venice: Institution of Engineering and Technology (IET), 2025, 167–173

[5]

Schick T, Dwivedi-Yu J, Dessi R, Raileanu R, Lomeli M, Hambro E, Zettlemoyer L, Cancedda N, Scialom T. Toolformer: language models can teach themselves to use tools. In: Advances in Neural Information Processing Systems 36. New Orleans: NeurIPS, 2023: 68539–68551

[6]

Yao S Y , Zhao J , Yu D , Du N , Shafran I , Narasimhan K , Cao Y . ReAct: synergizing reasoning and acting in language models. , 2022,

[7]

Schweidtmann A M . Generative artificial intelligence in chemical engineering. Nature Chemical Engineering, 2024, 1(3): 193

[8]

Daniel T , Xuan J . Responsible use of generative AI in chemical engineering. Digital Chemical Engineering, 2024, 12: 100168

[9]

Zhang D , Liu W , Tan Q , Chen J D , Yan H , Yan Y L , Li J T , Huang W R , Yue X Y , Ouyang W L . et al. ChemLLM: a chemical large language model. , 2024,

[10]

Zhao Z H , Ma D , Chen L , Sun L T , Li Z H , Xia Y , Chen B , Xu H S , Zhu Z C , Zhu S . et al. Developing ChemDFM as a large language foundation model for chemistry. Cell Reports Physical Science, 2025, 6(4): 102523

[11]

Zhou J B , Xu F Y , Chang Z J , Liu D P , Li L L , Cui J , Li Y , Li X , Qian L , Zhang Z X . et al. From lab to fab: a large language model for chemical engineering. Chinese Journal of Catalysis, 2025, 73: 159–173

[12]

Yang Y F , Shi R H , Li Z C , Jiang S , Lu B L , Zhao Q B , Yang Y , Zhao H . BatGPT-chem: a foundation large model for chemical engineering. Research, 2025, 8: 827

[13]

Rupprecht S , Gao Q H , Karia T , Schweidtmann A M . Multi-agent systems for chemical engineering: a review and perspective. Current Opinion in Chemical Engineering, 2026, 51: 101209

[14]

Bran A M , Cox S , Schilter O , Baldassari C , White A D , Schwaller P . Augmenting large language models with chemistry tools. Nature Machine Intelligence, 2024, 6(5): 525–535

[15]

Sun J Y , Zhou J B , Wang H B , Liu W , Yuan J , Wang Y , Xie T , Tan L , Zhang H L , Zou Y P . et al. TeLLAgent: a dual-agent framework for reliable scientific discovery with tool-enhanced LLMs. Chemical Science, 2026, 17(26): 13101–13117

[16]

Ding K Y , Yu J , Huang J J , Yang Y C , Zhang Q , Chen H J . SciToolAgent: a knowledge-graph-driven scientific agent for multitool integration. Nature Computational Science, 2025, 5(10): 962–972

[17]

Heyer M , Zhang J , Sugisawa N , Laub J F , Lapkin A A . Automated generation of mechanistic models for chemical process digital twins using reinforcement learning part I: conceptual framework and equation generation. Computers & Chemical Engineering, 2025, 202: 109281

[18]

Lee J , Park S , Oh S , Ma B. . Can large language models automate the HAZOP process without human intervention?. Safety Science, 2026, 194: 107039

[19]

Khan A , Nahar R , Chen H , Constante Flores G E , Li C . FaultExplainer: leveraging large language models for interpretable fault detection and diagnosis. Computers & Chemical Engineering, 2025, 199: 109152

[20]

Ren L , Wang H T , Wang Y Q , Huang K K , Wang L H , Li B H . Foundation models for the process industry: challenges and opportunities. Engineering, 2025, 52: 53–59

[21]

Bai L , Cai Z R , Cao Y H , Cao M S , Cao W H , Chen C Y , Chen H J , Chen K , Chen P C , Chen Y . et al. Intern-S1: a scientific multimodal foundation model. , 2025,

[22]

Bai L , Cao J Q , Chen C Y , Chen G Z , Chen K , Cheng G R , Cui E F , Dai X L , Ding S Y , Du S H . et al. Intern-S2-preview: scientific agentic foundation model. , 2026,

[23]

Bahamdan A , Pajak E , Hedengren J D , del Rio Chanona A . Sketch2Simulation: automating flowsheet generation via multi agent large language models. , 2026,

[24]

Srinivas S S , Gupta S , Runkana V . AutoChemSchematic AI: agentic physics-aware automation for chemical manufacturing scale-up. , 2025,

[25]

Zeng T , Badrinarayanan S , Ock J , Lai C K , Barati Farimani A . LLM-guided chemical process optimization with a multi-agent approach. Machine Learning: Science and Technology, 2025, 6(4): 045067

[26]

Yang Y H , Li R K , Ma J F , Zhang K , Liu Q , Han J Y , Bu Y G , Zhou J B , Lian D F , Li X . et al. CeProAgents: a hierarchical agents system for automated chemical process development. , 2026,

[27]

Schäfer P , Krinke L J , Wlotzka M , Asprion N . Context is all you need: toward autonomous model-based process design using agentic AI in flowsheet simulations. AIChE Journal, 2026, 72(10): e70488

[28]

Vyas J , Mercangoz M . Autonomous control leveraging LLMs: an agentic framework for next-generation industrial automation. , 2025,

[29]

Pajak E, Bahamdan A, Hellgardt K, Rio-Chanona A. Multi-agent LLMs for automating sustainable operational decision-making. In: Proceedings of the 35th European Symposium on Computer Aided Process Engineering (ESCAPE 35). Ghent: PSE Press, 2025: 1824–1829

[30]

Zheng Y Z , Koh H Y , Ju J X , Nguyen A T N , May L T , Webb G I , Pan S R . Large language models for scientific discovery in molecular property prediction. Nature Machine Intelligence, 2025, 7(3): 437–447

[31]

Frutiger J, Jones M, Ince N G, Sin G. From property uncertainties to process simulation uncertainties—Monte Carlo methods in SimSci PRO/II process simulator. In: 13th International Symposium on Process Systems Engineering (PSE 2018). Amsterdam: Elsevier, 2018: 1489–1494

[32]

Kennedy M C , O’Hagan A . Bayesian calibration of computer models. Journal of the Royal Statistical Society Series B: Statistical Methodology, 2001, 63(3): 425–464

[33]

Farquhar S , Kossen J , Kuhn L , Gal Y . Detecting hallucinations in large language models using semantic entropy. Nature, 2024, 630(8017): 625–630

[34]

Liu N F , Lin K , Hewitt J , Paranjape A , Bevilacqua M , Petroni F , Liang P . Lost in the middle: how language models use long contexts. Transactions of the Association for Computational Linguistics, 2024, 12: 157–173

[35]

Cemri M , Pan MZ , Yang S , Agrawal LA , Chopra B , Tiwari R , Keutzer K , Parameswaran A , Klein D , Ramchandran K . et al. Why do multi-agent LLM systems fail. In: Advances in Neural Information Processing Systems 38. San Diego: NeurIPS, 2025, 135676–135729

[36]

Li H T , Sarkar S , Lu W X , Loftus P O , Qiu T Y , Yu S E , Cuomo A E , Webster J P , Kelly H R , Manee V . et al. Collective intelligence for AI-assisted chemical synthesis. Nature, 2026, 651(8104): 107–115

[37]

Kim Y , Gu K , Park C , Park C , Schmidgall S , Heydari A A , Yan Y , Zhang Z H , Zhuang Y C , Liu Y . et al. Towards a science of scaling agent systems. , 2025,

[38]

Kapoor S , Narayanan A . Leakage and the reproducibility crisis in machine-learning-based science. Patterns, 2023, 4(9): 100804

[39]

Dwork C , Feldman V , Hardt M , Pitassi T , Reingold O , Roth A . The reusable holdout: preserving validity in adaptive data analysis. Science, 2015, 349(6248): 636–638

Rights & permissions

Higher Education Press

PDF (1058KB)

606

Accesses

0

Citation

Detail

Sections
Recommended

/

〈 〉