Tianming Liu’s research aims to transform infrastructure design by using AI to integrate and manage complex, conflicting human values in the engineering and design process. His work will develop a human-centered AI co-pilot framework using multi-agent Large Language Models that serve as calibrated proxies for complex systems and represent the interests of diverse stakeholders. By translating qualitative feedback into mathematical parameters and collaborating directly with engineers to negotiate design trade-offs, he aim to design critical systems, such as transportation and power grids, that are not just technically effective but also socially responsive.
