Events

​​AI-powered in-car voice assistants are rapidly evolving from reactive systems to proactive, adaptive companions. But how can this growing intelligence be evaluated in a systematic way – and how can future user experiences be tested early in the development process?

​In this talk, we explore the future design and evaluation of intelligent human–vehicle interaction using Large Language Models (LLMs) and AI agents from two perspectives:

​Part 1 focuses on the methodological evaluation of existing voice assistants: How can Large Language Models be used as “judges” to systematically benchmark the intelligence of current voice assistants? Where do today’s systems reach their limits?

​Part 2 addresses the simulation of future user experiences: Using AI-based personas and the concept of “human digital twins”, we show how complex driving contexts and diverse user expectations can be represented and tested early in the development process.

We will discuss these questions with you and our external expert Dr. rer. nat. Chandan Kumar (Team Leader „Interaction Design and Technologies“ am Fraunhofer IAO).

The event is aimed at

Research and Development in the Automotive and Mobility Sector: Software Development, AI Development, UI/UX Design and Research, HMI Development