Quality Engineering for GenAI-ALM Integration in Automotive Requirements Management

This thesis proposes a theoretical framework and a candidate architecture for the integration of Generative Artificial Intelligence (GenAI) within Application Lifecycle Management (ALM) platforms, to address the challenges, safety constraints and regulatory demands of the Requirements Management in the automotive industry. The research was conducted in collaboration with MCA Engineering S.r.l., an international engineering and high-tech consultancy company based in Turin, operating across numerous innovation-driven projects in the automotive sector. The preliminary phase comprises a comprehensive literature review and market analysis, followed by the identification of the product quality model. The analysis prioritizes four of the nine characteristics defined by the ISO/IEC 25010 standard, namely: maintainability, reliability, security, and usability. In order to quantify each attribute, Key Quality Indicators (KQIs) were derived according to the mathematical functions outlined in the ISO/IEC 25023 standards. Adopting a design science research methodology, the project led to the formulation of a conceptual proposal which includes a make-or-buy assessment, a project roadmap, and a risk evaluation through Failure Modes and Effects Analysis (FMEA). The conceptual framework and the logical architecture contributed, arise in response to the immaturity of the current market landscape and the evident academic gap in the GenAI-ALM integration in the automotive sector. The model illustrates the system’s core components, benefits, cost structure, and potential failure modes and outlines the key GenAI-powered functionalities such as requirement disambiguation, summarization of requirement documents, and translation of natural language inputs into technical specifications. The proposed solution will undergo a structured validation phase during the following internal deployment in MCA Engineering, through expert evaluation, to improve the system for future commercialization.