Understanding Software Quality Management in Automotive Software Development

Abstract

Managing software quality within the automotive context is of importance as the automotive industry is becoming increasingly software defined. Especially since the recalls of vehicles have increased due to software defects. Furthermore, the theory highlight challenges in defining what software quality is. This study thereby aims to increase the understanding of how software quality is defined within the automotive context, what quality management practices are important, and what factors impact software development and quality. To address this, a qualitative case study was conducted at a global software supplier to the automotive industry. The study primarily focuses on development processes rather than the quality specification of the final software product. The data collection was conducted through observations, documentation reviews, and semi-structured interviews with 28 participants. The study shows that software quality is perceived as multidimensional and context dependent, with emphasis on both technical and customer-experience aspects. The study identified important automotive software quality attributes such as safety, security, robustness, and reliability. The study identified important software quality management practices, consisting of reviews, gates, testing, traceability, and continuous improvement. However, the study also indicated ambiguities in the understanding and application of software quality assurance and control. This lack of a shared conceptual understanding creates difficulties in aligning responsibilities, communicating quality expectations and improving processes. Finally, five main factors were identified impacting software development and quality: process, requirements, system complexity & dependencies, time and testing. These factors are related to unclear requirements, fragmented processes, time pressure, and limited testing capabilities that were recurring in the findings.