A Counterfactual Safety Margin Perspective on the Scoring of Autonomous Vehicles' Riskiness (2308.01050v4)
Abstract: Autonomous Vehicles (AVs) promise a range of societal advantages, including broader access to mobility, reduced road accidents, and enhanced transportation efficiency. However, evaluating the risks linked to AVs is complex due to limited historical data and the swift progression of technology. This paper presents a data-driven framework for assessing the risk of different AVs' behaviors in various operational design domains (ODDs), based on counterfactual simulations of "misbehaving" road users. We propose the notion of counterfactual safety margin, which represents the minimum deviation from nominal behavior that could cause a collision. This methodology not only pinpoints the most critical scenarios but also quantifies the (relative) risk's frequency and severity concerning AVs. Importantly, we show that our approach is applicable even when the AV's behavioral policy remains undisclosed, through worst- and best-case analyses, benefiting external entities like regulators and risk evaluators. Our experimental outcomes demonstrate the correlation between the safety margin, the quality of the driving policy, and the ODD, shedding light on the relative risks of different AV providers. Overall, this work contributes to the safety assessment of AVs and addresses legislative and insurance concerns surrounding this burgeoning technology.
- N. Kalra and S. M. Paddock, “Driving to Safety: How Many Miles of Driving Would It Take to Demonstrate Autonomous Vehicle Reliability?” Tech. Rep.
- F. Favaró, L. Fraade-Blanar, S. Schnelle, T. Victor, M. Pena, J. Engstrom, J. Scanlon, K. Kusano, and D. Smith, “Building a Credible Case for Safety: Waymo’s Approach for the Determination of Absence of Unreasonable Risk,” 2023.
- T. Victor, K. Kusano, T. Gode, R. Chen, and M. S. Waymo, “Safety Performance of the Waymo Rider-Only Automated Driving System at One Million Miles,” 2023.
- J. M. Scanlon, K. D. Kusano, J. Engström, and T. V. Waymo, “Collision Avoidance Effectiveness of an Automated Driving System Using a Human Driver Behavior Reference Model in Reconstructed Fatal Collisions,” 2022.
- M. Schwall, T. Daniel, T. Victor, F. Favarò, and H. Hohnhold, “Waymo Public Road Safety Performance Data,” Tech. Rep., 2020.
- A. Coston, A. Mishler, E. H. Kennedy, and A. Chouldechova, “Counterfactual risk assessments, evaluation, and fairness,” in Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, vol. 27, no. 20. New York, NY, USA: ACM, 1 2020, pp. 582–593.
- M. Höfler, “Causal inference based on counterfactuals,” BMC Medical Research Methodology, vol. 5, no. 1, pp. 1–12, 9 2005.
- J. M. Scanlon, K. D. Kusano, T. Daniel, C. Alderson, A. Ogle, and T. Victor, “Waymo simulated driving behavior in reconstructed fatal crashes within an autonomous vehicle operating domain,” Accident Analysis & Prevention, vol. 163, p. 106454, 12 2021.
- M. Igl, D. Kim, A. Kuefler, P. Mougin, P. Shah, K. Shiarlis, D. Anguelov, M. Palatucci, B. White, and S. Whiteson, “Symphony: Learning Realistic and Diverse Agents for Autonomous Driving Simulation,” 5 2022.
- Z. Zhong, D. Rempe, D. Xu, Y. Chen, S. Veer, T. Che, B. Ray, and M. Pavone, “Guided Conditional Diffusion for Controllable Traffic Simulation.”
- P. Hart and A. Knoll, “Counterfactual Policy Evaluation for Decision-Making in Autonomous Driving.”
- D. Nishiyama, M. Y. Castro, S. Maruyama, S. Shiroshita, K. Hamzaoui, Y. Ouyang, G. Rosman, J. Decastro, K.-H. Lee, and A. Gaidon, “Discovering Avoidable Planner Failures of Autonomous Vehicles using Counterfactual Analysis in Behaviorally Diverse Simulation,” Tech. Rep., 2020.
- C. Voloshin, A. Verma, and Y. Yue, “Eventual Discounting Temporal Logic Counterfactual Experience Replay.”
- A. Lavaei, L. Di Lillo, A. Censi, and E. Frazzoli, “Formal Estimation of Collision Risks for Autonomous Vehicles: A Compositional Data-Driven Approach,” 12 2021.
- M. Althoff and O. Stursberg, “Reachability Analysis and its Application to the Safety Assessment of Autonomous Cars.”
- K. Leung, A. Bajcsy, E. Schmerling, and M. Pavone, “Towards Data-Driven Synthesis of Autonomous Vehicle Safety Concepts,” 7 2021.
- P. Schneider, M. Butz, C. Heinzemann, J. Oehlerking, and M. Woehrle, “Towards threat metric evaluation in complex urban scenarios,” in 2021 IEEE International Intelligent Transportation Systems Conference (ITSC). IEEE, 9 2021, pp. 1192–1198.
- L. Westhofen, C. Neurohr, T. Koopmann, M. Butz, B. Schütt, F. Utesch, B. Neurohr, C. Gutenkunst, and E. Böde, “Criticality Metrics for Automated Driving: A Review and Suitability Analysis of the State of the Art,” Archives of Computational Methods in Engineering, vol. 30, no. 1, pp. 1–35, 1 2023.
- D. Rempe, J. Philion, L. J. Guibas, S. Fidler, and O. Litany, “Generating Useful Accident-Prone Driving Scenarios via a Learned Traffic Prior.”
- T. Stewart, “Overview of Motor Vehicle Crashes in 2020,” Tech. Rep., 2022.
- J. D. States, “The Abbreviated and the Comprehensive Research Injury Scales,” in 13th Stapp Car Crash Conference, 2 1969.
- A. C. Malliaris, K. H. Digges, and J. H. Deblois, “Relationships between crash casualties and crash attributes,” in SAE Technical Papers, 1997.
- M. Althoff, M. Koschi, and S. Manzinger, “CommonRoad: Composable benchmarks for motion planning on roads,” in IEEE Intelligent Vehicles Symposium, Proceedings, 2017.
- A. Zanardi, “Driving Games common tools,” 2022. [Online]. Available: https://github.com/idsc-frazzoli/dg-commons
- K. Kreutz and J. Eggert, “Analysis of the Generalized Intelligent Driver Model (GIDM) for Uncontrolled Intersections,” in 2021 IEEE International Intelligent Transportation Systems Conference (ITSC). IEEE, 9 2021, pp. 3223–3230.
- Alessandro Zanardi (9 papers)
- Andrea Censi (34 papers)
- Margherita Atzei (2 papers)
- Luigi Di Lillo (6 papers)
- Emilio Frazzoli (100 papers)