Search Results

Machine Learning (ML) based technologies are increasingly being used to fulfill safety-critical functions in autonomous and advanced driver assistance systems (ADAS). This change has been spurred by recent developments in ML and Artificial Intelligence techniques as well as rapid growth of computing power. However, demonstrating that ML-based systems achieve the necessary level of safety integrity remains a challenge. Current research and development work focused on establishing safe operation of ML-based systems presents individual techniques that might be used to gain confidence in these systems. As a result, there is minimal guidance for supporting a safety standard such as ISO 26262 - Road Vehicles - Functional Safety, to enable the development of ML-based systems. This paper presents a survey of recent ML literature to identify techniques and methods that can contribute to meeting ISO 26262 requirements.

Over the past few decades, advanced methods have been developed for the analysis of digital systems using mathematical reasoning, i.e., formal logic. These methods are supported by sophisticated software tools that can be used to perform analysis far beyond what is practically achievable using “paper and pencil” analysis. In December 2011, RTCA published RTCA DO-178C [1] along with a set of technical supplements including RTCA DO-333 [2] which provides guidance on the use of formal methods towards the certification of airborne software. Such methods have the potential to reduce the cost of verification by using formal analysis instead of conventional test-based methods to produce a portion of the verification evidence required for certification.

The auto industry has had decades of experience with designing safe vehicles. The introduction of highly integrated features brings new challenges that require innovative adaptations of existing safety methodologies and perhaps even some completely new concepts. In this paper, we describe some of the new challenges that will be faced by all OEMs and suppliers. We also describe a set of generic top-level potential hazards that can be used as a starting point for the Preliminary Hazard Analysis (PHA) of a vehicle software-intensive motion control system. Based on our experience with the safety analysis of a system of this kind, we describe some general categories of hazard causes that are considered for software-intensive systems and can be used systematically in developing the PHA.