National Center for Sustainable Transportation

This project is to investigate how accessibility of city blocks is quantified through the transport systems and real traffic flow datafrom the Los Angeles Metropolitan Area. The authors investigate the reachability problem and provide a solution with a functional system that is capable of visualizing the reachability map (isochrone). Unlike other studies, this approach is data-driven and does not depend on mathematical graph-theory to compute the isochrone which requires intensive computation. Instead, it focuses on directly processing the large amount of traffic flow data that the Integrated Media Systems Center at USC has collected from the Regional Integration of Intelligent Transportation Systems (RIITS) for more than 10 years under the Center’s existing Archived Traffic Data Management System (ADMS) project. The reachability map construction is based on vehicle trajectories so the researchers devised the Data-Driven Trajectory Generator (DDTG), a data-driven, model-free, and parameter-less algorithm for generating realistic vehicle trajectory datasets from ADMS data. Since real world traffic is incomplete with lots of temporal and spatial missing data, the researchers studied imputation and interpolation methods to complete the dataset. Their experiments with real-world trajectory and traffic data show that the datasets generated by DDTG follow distributions that are very close to the distributions of a real trajectory dataset. Furthermore, to demonstrate the resultsfrom the proposed research, a web application was developed in which users can select a location, travel time, and the time of year to see the evaluated accessibility info in the form of an isochrone map. The outcomes of this project—synthetic vehicle trajectory dataset and reachability map construction—will be helpful in evaluating accessibility of city blocks for transport systems over a large area, essential for policymakers for effective city planning as well as to improve the well-being of citizens.

View the NCST Project Webpage