Data-Driven Synthesis of Hazardous Scenarios at Traffic Intersections


At a glance

Intersections are the most challenging elements of a transportation network in terms of safety as agents (human or robotic) face high complexity and uncertainties that can lead to wrong decisions and end in crashes. Our goal, in this project, is to automate synthesis of hazardous scenarios for urban intersections through a software toolkit. Our work addresses the following sponsor needs: 1) combining real-world data with simulation for the analysis of automated vehicles; 2) reliable motion prediction for traffic participants in a complex traffic environment, and 3) safe and robust decision making for automated driving in a complex driving environment.

To see the most current work built on results from this project, see "Data-Driven Generation and Analysis of Hazardous Traffic Scenarios".

principal investigatorsresearchersthemes
Sanjit Seshia
Pravin Varaiya
 traffic intersections, hazard analysis, control, simulation, verification, intelligent infrastructure