Travel forecasting provides essential insights into the movement patterns of people and goods within regions, playing a crucial role in guiding infrastructure development, informing policy decisions, and supporting strategic growth planning. In this talk, we will briefly review the transformative journey of travel forecasting methods over the past seven decades and examine the emerging role of artificial intelligence in this field, highlighting its potential to revolutionize model development. Our focus will be on advocating for a multi-scale modeling approach and exploring two AI-driven applications to transform this paradigm:

(1) an end-to-end learning approach that leverages extensive empirical traffic data to construct integrated network equilibrium models for light-duty policy analysis and prescription, and

(2) the application of large language models to enable generative agents to enhance activity-based microsimulation for in-depth, comprehensive analysis of plans and policies.

Dr. Yafeng Yin