Coastal erosion causes over $500 M in annual U.S. property losses, yet mitigation is a zero-sum challenge: inducing deposition in one area requires downcoast erosion. Effective, targeted interventions require a suite of structures tuned to different levels of deposition. This project develops 3D-printable concrete reef units generated via multi-objective Bayesian optimization in a nested-loop proposal-validation workflow. Physical testing in MIT’s 30 m wave tank replaces costly multiphase CFD, enabling broader exploration of the design space. The end goal is a portfolio of pareto-optimal mass-to-deposition reef geometries for precise, community-specific shoreline protection.

Growing up in Houston, especially in the wake of Harvey, coastal erosion was a highly visible problem. I’m excited to combine research tools, like tow tank testing, that I picked-up in previous UROPS, with coursework on optimization, that I got from classes like 2.086 and 2.155, to address a problem I’ve been interested in since high school.