(459f) Teaching Data-Driven Techniques to Chemical Engineering Majors

With the rapid technological advances in process industries, operational challenges also increase, requiring good understanding of chemical/physical principles, advanced control mechanisms, techno-economic evaluation criteria, as well as timely response to perceived anomalies, pushing the capabilities of any human operator to its limit. Benefiting from the re-emergence and recent broad application of artificial intelligence, data-driven methods, also combining the first-principle knowledge, have become a focus of attention in both industrial and academic communities. As students in the chemical engineering major require access to the current and ever-expanding AI techniques in research and industrial operation, topics in signal processing and pattern recognition will become as important as other courses in the curriculum. Yet, it is recognized that the chemical engineering curricula are already highly structured and demanding, introducing a new course sequence presents a substantial challenge, On the other hand, there are opportunities in the mathematics and engineering courses for students in the major to have a quick start in this area. A course designed particularly for seniors in chemical engineering can help students get ready for further improvement and application of AI in all areas of chemical engineering curriculum. The idea is to start from the knowledge of chemical process operations, and instrumentation and control courses to introduce the concept and value of data collection and processing, with the goal of sustaining safe and profitable operations, then bridge them with data-driven methods to face the challenges in modern industrial operations. With the well-developed and available open-source programming resources, one does not require every practitioner to become an expert in data science or an AI developer, instead, it will suffice to know how to rationally select the appropriate method and organize a workflow to address specific applications.

Based on this concept, we proposed a course with the title “Big Data and Chemical Engineering” for senior students in the major, and have piloted it as a free elective at Beijing University of Chemical Technology (BUCT) for three years. It includes data features from chemical processes, process monitoring beyond DCS platforms, the evaluation of process monitoring algorithms, open-source materials available for process monitoring algorithm integration, and industrial cases for process monitoring. Our data shows that it is well accepted by seniors, as the course builds on their mathematics, chemical engineering, and programming fundamentals. It will become a required course for seniors starting in Spring, 2025.