Search

The operational safety of Automated Driving System-equipped vehicles (AVs) is a critical issue with AVs being deployed on public roads. Methodologies for evaluating the operational safety are therefore necessary to maintain public safety. One possible approach is a safety case established by the AV developer that uses evidence to support a structured argument that the AV exhibits a given level of operational safety. One of the key components of a safety case for AVs is a set of testing results showing behavioral competency in a variety of scenarios within the AV’s operational design domain (ODD). The Institute of Automated Mobility (IAM) has previously published operational safety assessment (OSA) metrics along with a means to evaluate the severity of violations of the safety envelope-type OSA metrics for navigation of individual scenarios in the proposed OSA Methodology. The objective of the OSA Methodology is to objectively quantify the safety performance of each scenario navigated by the AV, with the aggregate quantifications to be included in the AV’s safety case. This work expands on the introduction of the OSA Methodology to (1) include more OSA metrics (beyond the safety envelope type) with clearly defined formulations for the severity of metric violation formulations, and (2) propose a methodology for interpreting the various OSA metric violation measurements to result in the desired quantification of the AV’s safety performance of a scenario navigation that is composed of evaluations of collision events, near-miss events, and non-collision events. The OSA Methodology is implemented in various simulation software and in real-world testing results at an IAM testbed to validate the concept. The results show a promising test result interpretation methodology that can be incorporated into a safety case framework to help ensure AV operational safety.

The automotive industry is transforming to a greater degree that has occurred since Henry Ford introduced mass production of the automobile with the Model T in 1913. Advances in computing, data processing, and artificial intelligence (deep learning in particular) are driving the development of new levels of automation that will impact all aspects of our lives including our vehicles. What are Connected and Automated Vehicles (CAVs)? What are the underlying technologies that need to mature and converge for them to be widely deployed? Fundamentals of Connected and Automated Vehicles is written to answer these questions, educating the reader with the information required to make informed predictions of how and when CAVs will impact their lives. Topics covered include: History of Connected and Automated Vehicles, Localization, Connectivity, Sensor and Actuator Hardware, Computer Vision, Sensor Fusion, Path Planning and Motion Control, Verification and Validation, and Outlook for future of CAVs.

Traffic monitoring cameras are powerful tools for traffic management and essential components of intelligent road infrastructure systems. In this paper, we present a vehicle localization and traffic scene reconstruction framework using these cameras, dubbed as CAROM, i.e., "CARs On the Map". CAROM processes traffic monitoring videos and converts them to anonymous data structures of vehicle type, 3D shape, position, and velocity for traffic scene reconstruction and replay. Through collaborating with a local department of transportation in the United States, we constructed a benchmarking dataset containing GPS data, roadside camera videos, and drone videos to validate the vehicle tracking results. On average, the localization error is approximately 0.8 m and 1.7 m within the range of 50 m and 120 m from the cameras, respectively., Comment: Accepted to IEEE ICRA 2021