2026 ASEE Annual Conference & Exposition

Generative Artificial Intelligence (GenAI) has been integrated into chemical process design engineering education at universities. This integration includes applications in control structure topologies, the development and autocompletion of process diagrams, and the generation of process flow diagrams, which are cross validated via process simulators, among others. While early evidence suggests promising learning outcomes and student engagement, evidence for its implementation effectiveness remains relatively limited. Most studies involve small cohorts and pilot-scale initiatives, which might not be representative enough to make sound conclusions. On the other hand, review papers highlight gaps in thorough research on GenAI in undergraduate engineering, as pedagogical frameworks for integrating these tools are still being conceptualized and integrated into engineering education. This work presents a systematic review of the use of GenAI in chemical process design undergraduate courses, focusing on clustering the specific domains of its use, as well as effectiveness assessment tools to meet learning outcomes and foster the mapping of Education 5.0 and Industry 5.0 elements. Moreover, we offer insights into a potential roadmap for GenAI integration into chemical engineering capstone design courses, covering the current development, deployment, assessment, and challenges of examples implemented in a third-year capstone course. In these examples, we explore strategies for "effective prompt engineering", design pipelines, and cross-validation of GenAI responses, which support the development of process deliverables, including process flow diagrams, piping and instrumentation diagrams, heat and mass balance, safety studies, circularity metrics, and economic evaluations of chemical processes. While this course focuses on team strategies for engineering design and the workflow is not entirely based on an open-ended project, it offers an earlier opportunity in our curriculum to guide students in using GenAI for chemical process design. This course is delivered in the semester right before the final capstone course in the fourth year of our program. Preliminary results reflect divergent levels of certainty and usefulness of GenAI tools among students when developing and/or validating process deliverables. Safety-related deliverables supported with Gen-AI are fairly described (according to the instructor and subject matter expert) because prompts are scaffolded with previous deliverables, including process diagrams and heat and material balance. However, process diagrams notably differ from standardized representations and, in some cases, lack well-structured control loops. Heat and material balances are efficiently prompted with process simulations, while deep troubleshooting is required to refine calculations. Effectiveness metrics for measuring GenAI prompt engineering and cross-validation are estimated and assessed by the subject matter expert. These metrics include comparing the deliverables against codes, standards, and regulations, as well as reference-based calculations developed in Microsoft Excel ® and process simulators. This work serves as guidance for engineering education researchers looking at a holistic and balanced approach to support students in developing process engineering deliverables while addressing ethical concerns associated with the integration of human-GenAI.