Authors: Camila Correa-Jullian, Ali Mosleh, and Jiaqi Ma.

About

Vehicle-to-Everything (V2X) communication paradigms have been developed to address safety challenges in the transportation environment. This combines vehicle and infrastructure technologies, relying on a combination of sensor data, wireless connectivity networks, and real-time calculations to support Advanced Driving Assistance Systems (ADAS) and Automated Driving Systems (ADS) applications. Despite the significant efforts made towards designing and implementing V2X messages, there is a need for systematic approaches to determine the safety impact and risk reduction benefits of V2X technologies, including key aspects of connectivity, hardware, software, and human reliability requirements. Hence, we seek to apply Probabilistic Risk Assessment frameworks to estimate the risk-reduction impact of V2X technologies, seeking to provide both qualitative and quantitative insights to risk management, even under limited data regimes.

This work presents a quantitative risk assessment methodology for V2X technologies based on the Hybrid-Causal Logic (HCL), a multi-layer framework that integrates event sequence diagrams, fault trees, and Bayesian networks providing a model-based approach to system analysis. This work draws from novel hazard identification methodologies, such as Concurrent Task Analysis and System-Theoretic Process Analysis, to define agents, critical events, and key tasks, failures, and errors, which are then modeled through HCL. The methodology incorporates the effect of driver-system cognitive team models based on Information, Decision, and Action in Crew context (IDAC) on the overall system’s safety, as well as enhancing context representation to consider effects of road types, weather conditions, traffic levels, and technology effectiveness under limited data availability through Bayesian networks.

Model-based risk assessment approaches are an important complement to simulation- and testing-based approaches used in transportation and mobility research to estimate the potential risk-reduction benefits of emerging technologies. Functional safety assessments have played a key role in establishing reliability, verification, and validation requirements in industry best practices. However, a more comprehensive approach to analyze the impact of interaction of human, hardware, and software on operational safety is required to support deployment decisions. Stakeholders at the local and state level require improved qualitative and quantitative methods to assess traffic and safety impacts of emerging technologies and would benefit from risk-informed decision support tools to develop resource allocation strategies. Further, model-based approaches can provide key input to determine data collection and testing priorities.

Scenario-based risk assessments can play an important role in contributing to the safer design and deployment of emerging technologies. As ADAS and ADS technology, regulation, and adoption evolve, it is likely that in the medium term the traffic ecosystem will exhibit a hybrid mix of vehicles at different Level of Driving Automation and Cooperative Driving Automation capabilities. However, the success of V2X technologies to improve traffic safety largely depends on the deployment of infrastructure-side technology – where the role of local and state stakeholder is key. This work presents an initial approach to develop a decision-making support tool for stakeholders to better understand the implications of V2X technologies applied at scale and the importance of pursuing data collection initiatives to reduce the uncertainty surrounding novel approaches.

Slides coming soon!