Fraunhofer Institute of Optronics, System Technologies and Image Exploitation IOSB 
Observe drivers, understand users
How do people behave in automated vehicles? When do they feel safe? And how does artificial intelligence recognize whether someone is distracted or overwhelmed? These and many other questions are answered by the new mobile research laboratory of the Human-AI Interaction (HAI) department at Fraunhofer IOSB - a test vehicle equipped with state-of-the-art sensors and AI technology at automation levels 2 and 3.
Any number of sensors can be installed in the vehicle interior to carry out all kinds of tests during real automated journeys.
The AktiMeter in the test vehicle reveals how ancillary activities are actually carried out during automated driving.
Complete measurement technology in the test vehicle. Edge-computing AI locally in the vehicle, data recording and cables to every seat.
The test vehicle: platform for adaptive human-AI interaction in the vehicle
Our test vehicle offers the following functionalities:
- Automation level 3 up to 95 km/h for realistic test situations in road traffic,
- Pedals on the passenger side for use by driving school staff - for safe testing on closed-off routes with heavily distracted drivers,
- Autarkic edge AI computing architecture (including NVIDIA Jetson/Orin) for executing sophisticated AI applications such as:
- Image processing
- Generative KI (LLM/VLM)
- Real-time capable AI agents for adaptive user interaction
- Flexible interior cabling for modular camera setups for monitoring driver and passenger behavior,
- Multimodal interaction interfaces via touch displays, combinable with generative AI,
- Sensor roof box with wipers, prepared for camera, lidar and radar for precise exterior visibility - even in wet or snowy conditions,
- Broad indoor and outdoor sensor portfolio for flexible use case development,
- AktiMeter-based interior analysis, for the detection and evaluation of secondary activities such as reading, sleeping or interaction with on-board displays,
- Expertise in user studies - from data collection and behavioral analysis to the acceptance assessment of new technologies,
Research for the mobility of tomorrow
The aim is to make the interaction between humans and AI adaptive, explainable and safe - especially in situations with shared control between driver and vehicle. The platform supports research and industry partners in:
- Benchmarks of driver und occupant monitoring systemen,
- the development of functions for interior cameras,
- the evaluation of HMI designs for indoor and outdoor communication,
- the detection of distraction, fatigue or motion sickness, and cognitive distraction,
- testing generative human-AI interfaces in vehicles,
- and the safe integration of new sensors and functions in realistic driving environments.