Infrastructure-Based Sensor Data Capture Systems for Measurement of Operational Safety Assessment (OSA) Metrics

Niraj Altekar, Steven Como, Duo Lu, Jeffrey Wishart, Donald Bruyere, Faisal Saleem, K. Larry Head

Research output: Contribution to journalConference articlepeer-review

Abstract

The operational safety of automated driving system (ADS)-equipped vehicles (AVs) needs to be quantified for an understanding of risk, requiring the measurement of parameters as they relate to AVs and human driven vehicles alike. In prior work by the Institute of Automated Mobility (IAM), operational safety metrics were introduced as part of an operational safety assessment (OSA) methodology that provide quantification of behavioral safety of AVs and human-driven vehicles as they interact with each other and other road users. To calculate OSA metrics, the data capture system must accurately and precisely determine position, velocity, acceleration, and geometrical relationships between various safety-critical traffic participants. The design of an infrastructure-based system that is intended to capture the data required for calculation of OSA metrics is addressed in this paper. The designed multi-modal sensor system includes a combination of traffic video cameras, vehicle-to-infrastructure (V2I) roadside units (RSUs), National Transportation Communications for Intelligent Transportation System Protocol (NTCIP)-compliant signal controllers streaming Signal Phase and Timing (SPAT) data, and Light Detection and Ranging (LIDAR) sensors. The system is contrasted with other design options to evaluate trade-offs between capability and cost. The designed data capture system was deployed at a SMARTDrive ProgramSM Test Bed intersection in Anthem, AZ that has been developed by the University of Arizona Transportation Research Institute (TRI) in cooperation with the Maricopa County Department of Transportation (MCDOT). The intersection is equipped with a sensor system that includes a fiber optic data transfer backbone to support the data transfer to a server at the MCDOT Traffic Management Center. A measurement uncertainty (MU) analysis has been conducted using experimental data to better understand the performance and reliability of the proposed sensor system design. The data capture system will enable the development and validation of a methodology to continuously measure OSA metrics by gaining rich information through fusion of multimodal data collected from available sources for safety assessment of the transportation system that includes AVs.

Original languageEnglish (US)
JournalSAE Technical Papers
Issue number2021
DOIs
StatePublished - Apr 6 2021
EventSAE 2021 WCX Digital Summit - Virtual, Online, United States
Duration: Apr 13 2021Apr 15 2021

Keywords

  • Data capture
  • Infrastructure
  • Sensors
  • Traffic safety assessment
  • Vehicle-to-infrastructure

ASJC Scopus subject areas

  • Automotive Engineering
  • Safety, Risk, Reliability and Quality
  • Pollution
  • Industrial and Manufacturing Engineering

Fingerprint Dive into the research topics of 'Infrastructure-Based Sensor Data Capture Systems for Measurement of Operational Safety Assessment (OSA) Metrics'. Together they form a unique fingerprint.

Cite this