Abstract:

This study investigates the integration of Artificial Intelligence (AI) tools into an introductory thermodynamics course for engineering students. The course is structured into four modules: two focus on fundamental physics concepts, one on mathematical foundations (including calculus and differential equations), and one on numerical methods using MATLAB. By combining theoretical learning with applied problem-solving, the curriculum bridges the gap between conceptual understanding and real-world engineering applications. Students engage with practical challenges, supported by AI-driven tools such as ChatGPT and Mathematica, to deepen their learning experience and enhance problem-solving abilities.

The main objective of this study is to evaluate how AI integration enhances students’ conceptual grasp of thermodynamics, assessed through pre- and post-course evaluations using the Thermodynamics Concept Inventory (TCI). Additionally, the research examines the impact of AI on student engagement, motivation, and the development of problem-solving skills in complex thermodynamic scenarios.

The course employs a flipped classroom model, where students first explore and solve problems using AI tools outside of class, guided by structured prompts for analysis and research. AI-generated solutions are then validated through online calculators. In-class sessions focus on discussion, concept reinforcement, and quizzes to assess understanding without AI assistance. High-complexity challenges—such as analyzing Stirling engines, the Seebeck and Peltier effects in cooling systems, and biodiesel production—provide practical applications of theoretical principles, fostering deeper understanding through hands-on experience. The course is delivered via Canvas, with assessments conducted through quizzes and WebAssign exercises.

Preliminary findings indicate increased student motivation and engagement, attributed to the lab-based, step-by-step approach that connects theory with practical, real-world applications. AI tools show promise in improving problem-solving skills and conceptual understanding, with feedback highlighting their role in enhancing topic exploration and refining solutions. The study demonstrates that AI, when integrated within a flipped classroom framework, can effectively reinforce theoretical knowledge, promote active learning, and increase student autonomy. These results suggest that AI tools serve as effective complements to traditional teaching methods, enhancing both student performance and the depth of learning in complex engineering topics. The study provides a framework for AI integration in STEM education, with implications for improving pedagogy and learning outcomes in physics and engineering courses.

Keywords:

Higher Education, STEM education, Challenge Based Learning, Artificial Intelligence Applications.