Project Details

[Return to Previous Page]

Designing an AI Tutor for Thermodynamics: Student Needs Assessment and Prototype

Company: PSU Department of Mechanical Engineering

Major(s):
Primary: ME

Non-Disclosure Agreement: NO

Intellectual Property: NO

Undergraduate engineering students increasingly turn to AI tools for help with their coursework, but instructors and curriculum designers still know very little about how, when, and why students use AI to support their learning, especially in challenging technical courses like Thermodynamics. This project, sponsored by three faculty members in Mechanical Engineering with support from the Leonhard Center, invites a capstone team to help fill that gap through student-facing needs assessment, design thinking, and low fidelity prototyping. The team will take on three connected charges. First, the team will use AI tools extensively throughout their own capstone work and document how they use them, what works, what does not, and where AI helps or hinders their design process. These reflections will inform faculty guidance on effective AI use in project-based engineering courses across the college. Second, the team will conduct a needs assessment with undergraduate students who are currently taking Thermodynamics. Using interviews, surveys, and observational methods, the team will identify where students struggle, when and how they currently use AI for support, and what a well-designed AI tutor could offer that existing tools do not. Third, based on what they learn, the team will design and build a low fidelity prototype of a personalized AI-supported Thermodynamics tutor and test it with a small group of student users. This project is well suited to students interested in user research, design thinking, and human-centered design. Prior experience with undergraduate Thermodynamics is welcome but not required. This project is part of a Penn State research study on AI in engineering education. Students who join this team will (a) conduct human-subjects research and will receive the required human-subjects research training to do so, and (b) also participate as subjects in the study by sharing their own reflections on their AI use throughout the capstone project. No work beyond the normal expectations of the capstone project will be required. By joining this project, students indicate their consent to participate. Participation as a research subject can be withdrawn at any time. Full details will be shared at the project kickoff, and questions are welcome to the research PI: Dr. Andrea Gregg, axg251@psu.edu. Deliverables: 1. A structured reflection report on the team's own use of AI throughout the capstone project, including recommendations for other project-based course instructors. 2. A needs assessment report summarizing findings from Thermodynamics students, including personas, key user needs, and design implications. 3. A low fidelity prototype of a personalized AI-supported Thermodynamics tutor, tested with a small group of student users.