Project: #95 Computer Aided Design for Safe Autonomous Vehicles Progress Report - Reporting Period Ending: Sept. 30, 2018 Principal Investigator: Rahul Mangharam Status: Active Start Date: July 1, 2017 End Date: June 30, 2021 Research Type: Applied Grant Type: Research Grant Program: FAST Act - Mobility National (2016 - 2022) Grant Cycle: 2017 Mobility21 UTC Progress Report (Last Updated: Nov. 30, 2018, 8:02 p.m.) % Project Completed to Date: 100 % Grant Award Expended: 100 % Match Expended & Document: 100 USDOT Requirements Accomplishments We have established a toolchain and safety benchmarks to effectively search for unsafe driving interactions and evaluate the end-to-end safety of autonomous vehicles. The toolchain, APEX, Autonomous vehicle Plan Verification and Execution, is used with popular perception, planning and control algorithms used in real AVs. This effort addresses a key challenge establishing a common safety criteria that is a major hurdle in wider deployment of AVs. It provides more structure in evaluating risk compared to the on-road testing-only approach used by auto companies. Impacts Wide dissemination: This work has been widely disseminated to DoT, regional transportation groups (Grater Valley Forge Transportation Authority, DVRPC, Penn DoT), AV companies (Nvidia, Qualcomm, Intel/MobilEye, General Motors R&D, Tesla), and in academic conferences and universities (Clemson University, SUNY Stony Brook, Washington University at St. Louis, UC Berkeley, Temple University, Lehigh University, Princeton University, NYU Tier-1 UTC etc. Conference organized: Penn Mobility21 PI, Mangharam, organized the 2018 Mobility21 conference on Moving America Forward: Next-Gen Truck Freight Transportation Summit on Oct 24-25, 2018 on the University of Pennsylvania campus. It was a very successful event with over 150 national transportation experts, including US DoT, Penn DoT, NY DoT, NJ DoT, several incumbent and startup companies and academics. Keynotes included FMCSA Administrator Ray Martinez and Penn DoT Secretary Leslie Richards. The summit discussed safety, electrification, automation, employment, market and policy issues in the new truck freight landscape. More details at http://www.seas.upenn.edu/mobility21 Other The project tools and benchmarks are released as open source for the community to use. It is compatible with mainstream machine learning software used by AVs in perception and planning. Outcomes New Partners Clemson University, NYU DoT Tier-1 UTC Issues The project was successfully completed.