Risk Assessment for Teleoperations of Level 4 Automated Driving Systems
Research Team: Jiaqi Ma (lead), Camila Correa Jullian, Marilia Ramos, and Xin Xia
UC Campus(es): UCLA
Problem Statement: The growing interest in Automated Driving Systems (ADS) has driven technological and regulatory advancements across various applications. High-autonomy vehicles, whether privately-owned or part of Mobility as a Service (MaaS) fleets, are poised to reshape the transportation landscape. However, as the technical, commercial, and regulatory landscape evolves, recent incident reports from testing and small-scale deployments suggest the need for a more focused approach to operational safety. This includes addressing issues like traffic disruptions and incident management procedures. Operational safety goes beyond the functional safety of ADS-equipped vehicles and encompasses activities such as service monitoring, dispatching, maintenance, incident response, staffing, training, and passenger support. It becomes a critical factor when scaling operations in vehicles without trained safety drivers, prompting questions about how vehicle manufacturers, ADS developers, and fleet operators can ensure safety before widespread commercialization and deployment.
Project Description: This research identifies key safety risks associated Level 4 ADS-equipped vehicle operation for fleets employed for Mobility as a Service (MaaS) applications. The research goes beyond assessing the functional safety of the ADS-equipped vehicles to explore the role of fleet operators in ensuring the operational safety of the vehicle fleets through remote driving assistance functions. Key responsibilities of the fleet operators are identified related to implementing risk reduction measures related to organizational management of change, training remote supervisors, ensuring suitable working conditions, enforcing vehicle connectivity and dispatching requirements, and coordinating incident mitigation procedures, training, tools, and work conditions. The research employs a hazard identification methodology that combines traditional and innovative methods to analyze risks involving human, software, and hardware systems; and identifies twenty hazard scenarios arising from system failures, human errors, and unsafe interactions during different operational phases. These scenarios are ranked based on their impact on safety and resource intensity, enabling fleet operators to make better decisions regarding resource allocation. By implementing these actions, fleet operators can prevent and mitigate safety hazards in the operation of ADS-equipped fleets though remote monitoring and driving assistance functions. The hazards and risk mitigation activities identified in this study may also improve the operational safety of passenger vehicles equipped with ADS technology as they become more widely deployed in future large-scale commercial operations.
Status: Completed
Budget: $10,000