Real-world data for AI-based applications. KARLI & IAT.

Real-world data for AI-based applications. KARLI & IAT.

The Institute of Human Factors and Technology Management (IAT) at the University of Stuttgart represents a central part of the empirical work in the KARLI project: It collects subjective and objective measurement data for AI-based applications.

At the Institute of Human Factors and Technology Management (Institut für Arbeitswissenschaft und Technologiemanagement – IAT) at the University of Stuttgart, we deal with the integrated planning, design, and optimization of innovative products and processes. While researching the area of human-technology interaction, one focus is the human-centered design of artificial intelligence, or “human-centered AI”. We explore the concepts and methods to AI-based applications designs for personalized, efficient and trustworthy user interaction, especially in the area of vehicle interaction.

The KARLI project offers us as IAT the chance to get one step closer to researching human-centered AI application. In addition to the conceptual and theoretical level, the project focuses on empirical work and the collection of real-world data. We are implementing this in collaboration with our partner institute Fraunhofer IAO and their Wizard-of-Oz vehicle, a vehicle equipped with sensor technology for simulating various automation stages.

Our expertise in the project covers a wide range of surveys with subjective and objective measurement data. We design new user-adaptive interaction concepts for intelligent voice assistants in iterative user-oriented development steps. We then evaluate them with potential users. We also investigate potential consequences of automated driving, such as motion sickness, and their effects on the user in more detail under real-world conditions.

In KARLI, we primarily contribute the methodological expertise and the psychological and linguistic know-how for AI-based voice dialogues, which are indispensable for the application developments of the project partners Audi and Ford.

Our team at IAT is led by Lesley-Ann Mathis and Daniela Piechnik, who both bring an interdisciplinary background in cognitive science to the project. This research field combines many disciplines relevant to KARLI, e.g. linguistics, neuroscience, artificial intelligence and psychology.

Daniela brings her expertise in the field of physiological data acquisition and processing. This includes eye-tracking data to analyze eye movements or heart rate and skin conductivity to draw conclusions about the user’s current state. In regard to motion sickness, this brings out the technology to detect any changes in the driver’s well-being at an early stage.

Lesley-Ann is using her linguistics background to explore proactive voice interaction with a smart vehicle. In collaboration with Audi, she dives deeper into how to provide personalized support to users at the right time during their journey to provide an optimal user experience.

Additionally, we are also using this approach of proactive voice interaction to develop a self-rating tool for the early detection of motion sickness.

We are already looking forward to the first data collection sessions with users, in order to expand our methodological knowledge for human-centered design of AI and KARLI applications.

Leave a comment