Extraction of Road Users' Behavior From Realistic Data According to Assumptions in Safety-Related Models for Automated Driving Systems

TL;DR

The paper tackles deriving reasonably foreseeable behaviors of road users for automated driving systems by applying IEEE Std 2846-2022 to real-data. It defines four high-level scenarios (S1–S4) and extracts kinematic bounds from the UniD dataset using a Python-based pipeline to form safety envelopes around the ego vehicle. The contributions include a data-driven method to bound $v^{lon}(t)$, $v^{lat}(t)$, $α^{lon}(t)$, $α^{lat}(t)$, $β^{lon}(t)$, $β^{lat}(t)$, $h(t)$, $h'(t)$, and $λ(t)$ for different road users and scenarios. The results provide practical initial-condition and safety-bound parameters for ADS testing in simulation and on-road trials, enabling more informed planning under reasonably foreseeable risk. The work also outlines future extensions to richer scenarios and datasets.

Abstract

In this work, we utilized the methodology outlined in the IEEE Standard 2846-2022 for "Assumptions in Safety-Related Models for Automated Driving Systems" to extract information on the behavior of other road users in driving scenarios. This method includes defining high-level scenarios, determining kinematic characteristics, evaluating safety relevance, and making assumptions on reasonably predictable behaviors. The assumptions were expressed as kinematic bounds. The numerical values for these bounds were extracted using Python scripts to process realistic data from the UniD dataset. The resulting information enables Automated Driving Systems designers to specify the parameters and limits of a road user's state in a specific scenario. This information can be utilized to establish starting conditions for testing a vehicle that is equipped with an Automated Driving System in simulations or on actual roads.

Figures (5)

• Figure 1: Scenario S1: The ego vehicle (red) is driving next to another road user (a pedestrian in (a) or a blue car in (b)).
• Figure 2: Scenario S2: The ego vehicle (red) is driving behind another road user (a blue car in this example).
• Figure 3: Scenario S3: The ego vehicle (red) is driving behind a road user (blue) and in front of another (grey). In this example both road users are cars.
• Figure 4: Scenario S4: The ego vehicle (red) is driving while a (blue pedestrian) is crossing the road.
• Figure 5: Selected time frame of the UniD dataset depicting Pedestrians (red), cyclists (yellow), and vehicles (blue). Considering "ID78" as the ego vehicle, then there is one occurrence of scenario S1 with the cyclist "ID79" and one occurrence of scenario S4 with the pedestrian "ID45".