2025 ASEE Annual Conference & Exposition

Data-Driven Research Experience for Undergraduate Students

Abstract
Data analysis is essential to modern engineering systems and processes. With advanced computational tools, large datasets can be stored, processed, and analyzed to uncover key characteristics and trends. Developing the ability to make data-driven inferences and predictions is a crucial skill for today’s engineering students. This paper discusses the integration of innovative Artificial Intelligence (AI) deep learning techniques into undergraduate research at a teaching-focused institution, with an emphasis on data mining and infrastructure performance prediction.
The research projects highlighted in this paper involve analyzing Connecticut’s bridge inventory and inspection data for forecasting bridge conditions, as well as examining driver behavior, including speed profiles and gap acceptance at roundabouts in Connecticut. These projects are either internally and externally funded, providing students with opportunities to engage in research during the summer, throughout the academic year, or as part of their independent study courses. Some key student research tasks in these projects include data collection, data cleaning, contextualization, analysis using AI neural networks, and result interpretation and presentation. Students are given multiple opportunities to present their work, including at the University’s undergraduate research colloquium, progress meetings with Connecticut Department of Transportation staff, and at local and regional conferences.
This paper uses these projects as case studies to demonstrate how data-driven research experiences can be successfully integrated into undergraduate education, particularly at teaching-focused institutions without Ph.D. students. A survey was conducted among both current students and alumni who participated in these research projects, evaluating the impact of the experience on their technical skills, self-confidence, and ability to work independently. The survey results indicate that students reported significant improvements in time management, research skills, and self-confidence. These findings reinforce the hypothesis that data-driven research experiences contribute to both academic growth and the development of lifelong learning skills. In conclusion, this paper provides insights into how undergraduate students at primarily teaching institutions can benefit from hands-on, innovative data-driven AI research experiences. It also offers potential strategies for incorporating these valuable skills into future undergraduate research opportunities, enhancing student engagement and learning outcomes.