Deep Learning Safety Concerns in Automated Driving Perception

TL;DR

This paper discusses safety concerns related to deep learning perception in automated driving, extending existing frameworks with new categorizations to better address safety issues in AI-driven systems.

Contribution

It refines and extends the concept of safety concerns for DNNs in automated driving, incorporating expert feedback and new categorization for improved safety management.

Findings

Enhanced safety concern categorization for DNN perception

Alignment with safety standards like ISO 21448 and ISO PAS 8800

Facilitates cross-functional safety team collaboration

Abstract

Recent advances in the field of deep learning and impressive performance of deep neural networks (DNNs) for perception have resulted in an increased demand for their use in automated driving (AD) systems. The safety of such systems is of utmost importance and thus requires to consider the unique properties of DNNs. In order to achieve safety of AD systems with DNN-based perception components in a systematic and comprehensive approach, so-called safety concerns have been introduced as a suitable structuring element. On the one hand, the concept of safety concerns is -- by design -- well aligned to existing standards relevant for safety of AD systems such as ISO 21448 (SOTIF). On the other hand, it has already inspired several academic publications and upcoming standards on AI safety such as ISO PAS 8800. While the concept of safety concerns has been previously introduced, this paper extends and refines it, leveraging feedback from various domain and safety experts in the field. In particular, this paper introduces an additional categorization for a better understanding as well as enabling cross-functional teams to jointly address the concerns.