On June 12, 2026, a new paper titled 'TrajGenAgent: A Hierarchical LLM Agent for Human Mobility Trajectory Generation' was published on ArXiv AI.
This approach seeks to tackle the difficulties associated with collecting human mobility data, which is often hindered by costs and privacy issues.
By employing a hierarchical large language model, TrajGenAgent aims to produce realistic mobility trajectories that could benefit fields such as transportation, urban planning, and epidemic control.