Risk Analysis of Autonomous Vehicles in Mixed Traffic Streams

Dey, Kakan Chandra; Chowdhury, Mashrur (Ronnie); Das, Plaban · 2017 · ROSA P / City College of New York. University Transportation Research Center

archive: archived pipeline: cataloged verified

Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)

Summary

This study addresses the safety risks associated with autonomous vehicles (AVs) operating in mixed traffic streams, where they share roadways with human-driven vehicles. The research was motivated by the need to evaluate potential failure modes before mass implementation, particularly given that human error accounts for 94% of U.S. road crashes and AVs rely on complex sensors, actuators, and communication platforms vulnerable to disruption, hacking, or infrastructure issues. The primary objectives were to identify hierarchical sequences of events leading to AV failure, develop strategies to minimize these risks, and perform a benefit-cost analysis to determine the most economical mitigation measures. The researchers employed a three-phase risk analysis methodology: risk identification, risk estimation, and evaluation. To identify risks, the AV system was disassembled into vehicular components (sensors, actuators, communication platforms) and transportation infrastructure components. Failure probabilities for each component were estimated through literature reviews and validated via an online Delphi survey of experts. Fault tree models were developed for both vehicular and infrastructure failures to determine the likelihood of system-wide failure. These models were combined to assess overall risk in mixed traffic scenarios. The results were validated using real-world data from California Department of Motor Vehicles autonomous vehicle testing records. The fault tree analysis revealed that the probability of autonomous vehicle failure due to sequential vehicular component failures is approximately 14% over the vehicle’s lifetime. When combining failures from both vehicular and infrastructure components, the study determined an overall failure probability of 158 incidents per 1 million miles of travel. The analysis identified and ranked minimal cut-sets, which are the most critical combinations of events leading to failure. Based on these findings, the authors identified 22 specific strategies to minimize AV failure probabilities. These strategies were subsequently evaluated using benefit-cost analysis to assess their economic viability under different market penetration scenarios (10% in 2030 and 100% in 2050). The significance of this work lies in providing a quantitative framework for assessing AV safety during the transition phase of mixed traffic. By identifying specific failure probabilities and critical event combinations, the study offers actionable insights for improving AV reliability. The benefit-cost analysis of the 22 mitigation strategies provides decision-makers with evidence-based recommendations for implementing safety measures that balance cost and risk reduction. This research supports the development of safer, more reliable autonomous vehicles by highlighting the importance of addressing both vehicular component vulnerabilities and infrastructure-related disruptions.

Key finding

The fault tree analysis determined an overall failure probability of 158 incidents per one million miles of travel for autonomous vehicles in mixed traffic streams.

Methodology

modeling

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).

Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• driverless ads
• situational awareness
• pre crash contributing factors

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: crash risk outcomes