Guidelines for Designing User-Centered AI Agents for the Automotive Industry, Under the Hood of Creating an Interface to Support Perceived Quality Evaluation in Automotive Design

Abstract

This thesis investigates the design of user-centered AI agents for the automotive industry with a focus on their application in Perceived Quality (PQ) evaluation support. It addresses the challenges faced by automotive designers, engineers, and decision-makers by exploring their workflows, needs, and barriers to AI adoption. Using a human-centered design approach, the research incorporates qualitative in terviews and thematic analysis to understand current design processes and identify key concerns about integrating AI. The findings underscore the importance of trust, transparency, and collaboration, as well as the necessity of AI serving as a supportive tool rather than a replacement for human creativity. Based on these insights, nine actionable design guidelines are proposed to ensure that user-centered AI agents complement the workflows of decision-makers and designers while preserving their creative autonomy. Additionally, a high-fidelity prototype of a Perceived Quality AI agent was developed to demonstrate these guidelines in practice. This study contributes to the emerging field of human-centered AI design in industrial con texts. It also provides a foundation for future research to refine and validate the proposed design principles through user testing and further application not only in the automotive sector but also across other industries.