A Method for Scenario Risk Quantification for Automated Driving Systems

Recent innovations, such as automated driving and smart mobility, have elevated the safety-criticality of automotive systems due to the impact of these technologies on the traffic behavior and safety. New safety validation and assessment methodologies are required to provide the level of assurance that matches the societal impact of these systems. The objective of this paper is to introduce a novel method for assessment and quantification of the risk of a driving scenario considering the operational design domain. For the proposed method, the authors assume that a scenario consists of activities (performed by different actors) and environmental conditions that leads to a potentially hazardous consequence. The risk of a driving scenario is the product of the probability of the exposure of a scenario and the severity of the hazardous consequence of that scenario. The authors introduce a systematic method for calculating the probability of exposure, where they assume a causal relation between the activities that constitute a scenario. By making educated assumptions on the dependencies among the different activities and environmental conditions, the authors simplify the calculation of the probability of the exposure. For estimating the severity, they employ Monte Carlo simulations. They illustrate the use of their proposed method by applying it to an example of a collision avoidance system in a cut-in scenario. The authors use naturalistic driving data acquired from field studies on the Dutch highways to determine the risk. The presented example illustrates the potential of the proposed risk estimation method. Using the proposed method, the authors can compare the safety criticality of various scenarios in a quantitative manner, which can be used as a safety metric for evaluating automated driving systems. This can lead to stronger justification for design decisions and test coverage for developing automated vehicle functionalities.

• Record URL:
  https://www-nrd.nhtsa.dot.gov/departments/esv/26th/
• Record URL:
  https://www-esv.nhtsa.dot.gov/Proceedings/26/26ESV-000129.pdf
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/oclc/19-0051
• Supplemental Notes:
  • Paper Number: 19-0129-O
• Corporate Authors:

  National Highway Traffic Safety Administration

  1200 New Jersey Avenue, SE
  Washington, DC  United States  20590
• Authors:
  • de Gelder, Erwin
  • Saberi, Arash Khabbaz
  • Elrofai, Hala
• Conference:
  • 26th International Technical Conference on the Enhanced Safety of Vehicles (ESV): Technology: Enabling a Safer Tomorrow
  • Location: Eindhoven , Netherlands
  • Date: 2019-6-10 to 2019-6-13
• Publication Date: 2019

Language

• English

Media Info

• Media Type: Digital/other
• Features: Figures; References; Tables;
• Pagination: 11p
• Monograph Title: 26th International Technical Conference on the Enhanced Safety of Vehicles (ESV): Enabling a Safer Tomorrow

Subject/Index Terms

• TRT Terms: Behavior; Intelligent vehicles; Risk analysis; Traffic safety
• Geographic Terms: Netherlands
• Subject Areas: Highways; Safety and Human Factors;

Filing Info

• Accession Number: 01761603
• Record Type: Publication
• Report/Paper Numbers: 19-0129
• Files: TRIS, ATRI, USDOT
• Created Date: Dec 9 2020 3:28PM