Identification of Challenging Highway-Scenarios for the Safety Validation of Automated Vehicles Based on Real Driving Data

For a successful market launch of automated vehicles (AVs), proof of their\nsafety is essential. Due to the open parameter space, an infinite number of\ntraffic situations can occur, which makes the proof of safety an unsolved\nproblem. With the so-called scenario-based approach, all relevant test\nscenarios must be identified. This paper introduces an approach that finds\nparticularly challenging scenarios from real driving data (\\RDDwo) and assesses\ntheir difficulty using a novel metric. Starting from the highD data, scenarios\nare extracted using a hierarchical clustering approach and then assigned to one\nof nine pre-defined functional scenarios using rule-based classification. The\nspecial feature of the subsequent evaluation of the concrete scenarios is that\nit is independent of the performance of the test vehicle and therefore valid\nfor all AVs. Previous evaluation metrics are often based on the criticality of\nthe scenario, which is, however, dependent on the behavior of the test vehicle\nand is therefore only conditionally suitable for finding "good" test cases in\nadvance. The results show that with this new approach a reduced number of\nparticularly challenging test scenarios can be derived.\n