REU Site: Integrating Machine Learning and Causal Principles for Civil and Environmental Engineers

About This Grant

Modern infrastructure faces unprecedented challenges due to urbanization and aging systems, which require innovative solutions that traditional engineering approaches alone cannot provide. This Research Experience for Undergraduates (REU) Site at Clemson University addresses a critical national need by training the next generation of civil and environmental engineers (CEE) to harness the power of machine learning (ML) and causal inference. The program will immerse 12 undergraduate students annually in an 8-week intensive research experience that tackles real-world infrastructure problems such as ensuring equitable water distribution, optimizing transportation networks to reduce congestion and emissions, and designing infrastructure systems that can withstand natural disasters. Industry partnerships will ensure that research outcomes translate into practical solutions, while outreach activities will inspire younger students to pursue STEM careers. This program supports NSF's mission to advance national prosperity and welfare through improved infrastructure resilience and efficiency by bridging the gap between traditional engineering education and cutting-edge computational methods. This REU Site will integrate machine learning and causal inference principles into civil and environmental engineering research through mentored projects spanning multiple CEE domains. The program's technical objectives include: (1) developing novel ML methodologies for infrastructure optimization, including deep learning models for traffic flow prediction and reinforcement learning for adaptive infrastructure management; (2) creating causal inference frameworks to understand complex relationships between infrastructure performance and environmental factors; (3) building open-source ML tools and curated datasets specific to CEE applications; and (4) establishing ethical guidelines for ML deployment in infrastructure projects. The research methodology combines computational modeling, data analytics, and validation through partnerships with local infrastructure firms and agencies. Each project will incorporate modules on responsible ML development to ensure that participants understand the ethical implications of algorithmic decision-making in public infrastructure. The program will produce peer-reviewed publications, open-source software tools, and a pedagogical framework for integrating ML into undergraduate CEE curricula. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.