An Algorithmic Approach to Matching Students with Capstone Design Projects

Capstone design projects represent a prime element to the engineering education system.These projects require students to apply their accumulated knowledge to design innovative products or processes, simulating real-world engineering practices.Teams are typically formed to reflect the collaborative nature of engineering design firms.In many US universities, capstone projects receive funding either from the academic department or through partnerships with industry.Traditionally, students select projects based on personal interest, which can lead to challenges such as imbalanced teams, mismatches between student skills and project requirements, and potential loss of sponsorship due to poor project outcomes.To address these issues, we introduce an artificial intelligence-based algorithm that optimizes the matching of students to projects.The algorithm considers factors such as GPA, demographics, engineering interests, and relevant skills, and it continuously refines its process by analyzing team performance data.The implementation of this algorithm at the University of North Texas (UNT) has significantly improved student and sponsor satisfaction, with average satisfaction scores rising to 90%.Additionally, statistical analysis has shown a significant improvement (𝑝 < 0.0018) in overall summative rating before and after the algorithm's implementation.Sponsor retention has also increased, with 65% of sponsors returning each semester with new projects.This algorithmic approach ensures a fair, efficient, and effective team formation process, closely reflecting realworld engineering environments.