Context is all you need: Towards autonomous model-based process design using agentic AI in flowsheet simulations
arXiv:2603.12813v1 Announce Type: new Abstract: Agentic AI systems integrating large language models (LLMs) with reasoning and tooluse capabilities are transforming various domains - in particular, software development. In contrast, their application in chemical process flowsheet modelling remains largely unexplored. In this work, we present an agentic AI framework that delivers assistance in an industrial flowsheet simulation environment. To this end, we show the capabilities of GitHub Copilot (GitHub, Inc., 2026), when using state-of-the-art LLMs, such as Claude Opus 4.6 (Anthropic, PBC, 2026), to generate valid syntax for our in-house process modelling tool Chemasim using the technical documentation and a few commented examples as context. Based on this, we develop a multi-agent system that decomposes process development tasks with one agent solving the abstract problem using engineering knowledge and another agent implementing the solution as Chemasim code. We demonstrate the effe
arXiv:2603.12813v1 Announce Type: new Abstract: Agentic AI systems integrating large language models (LLMs) with reasoning and tooluse capabilities are transforming various domains - in particular, software development. In contrast, their application in chemical process flowsheet modelling remains largely unexplored. In this work, we present an agentic AI framework that delivers assistance in an industrial flowsheet simulation environment. To this end, we show the capabilities of GitHub Copilot (GitHub, Inc., 2026), when using state-of-the-art LLMs, such as Claude Opus 4.6 (Anthropic, PBC, 2026), to generate valid syntax for our in-house process modelling tool Chemasim using the technical documentation and a few commented examples as context. Based on this, we develop a multi-agent system that decomposes process development tasks with one agent solving the abstract problem using engineering knowledge and another agent implementing the solution as Chemasim code. We demonstrate the effectiveness of our framework for typical flowsheet modelling examples, including (i) a reaction/separation process, (ii) a pressure-swing distillation, and (iii) a heteroazeotropic distillation including entrainer selection. Along these lines, we discuss current limitations of the framework and outline future research directions to further enhance its capabilities.
Executive Summary
This article introduces an agentic AI framework for flowsheet simulation in chemical process design, leveraging GitHub Copilot and state-of-the-art large language models. The framework decomposes process development tasks into abstract problem-solving and code implementation, demonstrating effectiveness in three typical flowsheet modelling examples. While the framework shows promise, limitations and future research directions are also discussed. The work has implications for industry and academia, presenting a potential paradigm shift in process design, and highlights the need for further investigation into the integration of AI and process engineering.
Key Points
- ▸ Introduction of an agentic AI framework for flowsheet simulation
- ▸ Use of GitHub Copilot and state-of-the-art large language models
- ▸ Decomposition of process development tasks into abstract problem-solving and code implementation
Merits
Enhanced Efficiency
The framework has the potential to significantly enhance the efficiency of process design, reducing the time and effort required to develop and implement new processes.
Improved Accuracy
By leveraging large language models and GitHub Copilot, the framework can provide more accurate and reliable results, reducing the likelihood of errors and improving overall process quality.
Demerits
Limited Generalizability
The framework's effectiveness is demonstrated in a limited set of typical flowsheet modelling examples, and further research is needed to determine its generalizability to more complex and diverse process design scenarios.
Dependence on Technical Documentation
The framework's reliance on technical documentation and commented examples as context may limit its applicability to industries or scenarios where such documentation is lacking or incomplete.
Expert Commentary
The article presents a promising and innovative approach to flowsheet simulation in chemical process design, leveraging the capabilities of agentic AI and large language models. While there are limitations and challenges to be addressed, the framework has the potential to transform the field of process design and improve the efficiency and accuracy of process development. Further research is needed to explore the generalizability and applicability of the framework to diverse process design scenarios and to investigate the integration of AI and process engineering in more depth.
Recommendations
- ✓ Further investigation into the integration of AI and process engineering, including the development of more advanced frameworks and methodologies.
- ✓ Development of more comprehensive and accessible technical documentation and resources to support the widespread adoption of the framework.