Towards Efficient Hazard Identification in the Concept Phase of Driverless Vehicle Development

TL;DR

Hazard identification for driverless vehicle development faces a combinatorial explosion when coupling malfunction-driven behaviors with operational scenarios in the concept phase. The authors propose a deviation-based hazard identification approach that replaces exhaustive malfunction-scenario mapping with deviations from the desired externally observable vehicle behavior within predefined scenarios, formalizing the idea with $M$, $D$, and $S$ to compare $|P_M|=|M\times S|$ and $|P_D|=|D\times S|$ and showing $|P_M|>|P_D|$ while $f(P_M)\subseteq f(P_D)$. The method is demonstrated in the UNICARagil project, reducing redundant hazards and improving traceability by focusing on observable vehicle motions and preventing multiple malfunctions from generating the same hazard. The work contrasts the proposed approach with ISO 26262-based and expert-based strategies, highlighting efficiency gains in early design while acknowledging remaining challenges in scenario management and goal translation. Overall, deviations-based hazard identification offers a practical pathway to scalable, early-safety concepts for automated driving systems, with future work on systematic scenario specification and deriving component-level safety requirements from vehicle-level safety goals.

Abstract

The complex functional structure of driverless vehicles induces a multitude of potential malfunctions. Established approaches for a systematic hazard identification generate individual potentially hazardous scenarios for each identified malfunction. This leads to inefficiencies in a purely expert-based hazard analysis process, as each of the many scenarios has to be examined individually. In this contribution, we propose an adaptation of the strategy for hazard identification for the development of automated vehicles. Instead of focusing on malfunctions, we base our process on deviations from desired vehicle behavior in selected operational scenarios analyzed in the concept phase. By evaluating externally observable deviations from a desired behavior, we encapsulate individual malfunctions and reduce the amount of generated potentially hazardous scenarios. After introducing our hazard identification strategy, we illustrate its application on one of the operational scenarios used in the research project UNICAR$agil$.

Figures (3)

• Figure 1: Operational scenario of oncoming traffic on a two-lane road
• Figure 2: Illustration of an operational reference scenario in UNICARagil. Pedestrian P enters the road from an occlusion and halts for the automated ego-vehicle to pass. The ego-vehicle passes the waiting pedestrian with adjusted speed and lateral position.
• Figure 3: Stages of the discussed strategies for hazard identification. White boxes indicate the actual process steps for hazard identification, those with bold outlines mark the source of hazardous scenarios.