Mobility Data Landscape: Review, Fusion, and Synthesis for Transportation Insights

Problem Statement

Transportation agencies and researchers struggle with fragmented, incomplete, or unavailable mobility data, making it difficult to accurately model mobility patterns and predict future transportation demand. While various datasets exist—such as GPS trajectories, public transit records, traffic sensors, and household travel surveys—they are often disconnected, limited in scope, or proprietary, preventing cities from making fully informed planning decisions. These datasets, when properly integrated, have the potential to improve urban planning, transportation optimization, and system operations. This project aims to review available mobility data sources critical for mobility pattern analysis, build a mobility data fusion pipeline by using multiple cross-domain data sources, allowing for detailed synthesis and modeling of urban and rural activities, travel behavior, demand, and trajectories, as well as estimation/generation of network-wide travel patterns. Ultimately, this project will provide a scalable, transferable data fusion framework that agencies can use for demand prediction, transportation planning, and network optimization.