Project
KARLI
RESEARCH-
PROJECT
KARLI | Research Project for Level-Compliant Driver Behavior
PROJECT DESCRIPTION
Shaping
Smart
Mobility
The KARLI project, together with Ford, Audi, Continental, and Fraunhofer Institutes, researched the future of human-machine interaction in cars. A fascinating user experience study examined how understandable icons are and how smoothly the transition of driving control to artificial intelligence (AI) works. Funded by the German Federal Ministry for Economic Affairs and Climate Action, KARLI aims to make the mobility of tomorrow more accessible through reliable and user-friendly AI.
Vision
Adaptive
Driving
Assistance
The KARLI project improves future vehicle interaction by developing advanced AI functions. These functions precisely capture driver conditions and adjust to different levels of automation. The AI models use both real and synthetic data and are scalable for the vast amounts of data in future production vehicles. KARLI is paving the way for a new era of smart and safe vehicle technology.
Personas
Human-
Machine
Interaction
Together, 12 consortium partners analyze how automated vehicles can best meet user needs and where human-machine interaction should be improved. Personas were created, and their detailed user journeys were explored. An example: How can Lisa's winding commute to work be made more comfortable, and how can travel sickness be avoided? The goal: to develop future-oriented, user-friendly vehicle solutions based on these insights.
USER STUDY 1.0
Research
Immersive {{br}
Virtual
Reality
The user study took place in an immersive virtual reality (VR) environment equipped with a special chair, steering wheel, pedals, and the Unity engine. Twenty participants of different genders and ages experienced a realistic driving simulation. Through VR headsets, video data was transmitted to simulate various driving scenarios, from relaxed everyday drives to emergency situations. The questions asked during these scenarios were key to optimizing the driving experience.
RESULTS
Preferences
and
clearness
The study showed that a robotic design appealed to most users, especially younger and AI-savvy individuals. Older participants preferred the humanoid design. The robotic design scored high for comprehensibility (45%) and appealing aesthetics, while the humanoid design was often seen as unsettling. Abstract designs offered a good mix of clarity and functionality.
USER STUDY 2.0
Handover
Design
Evaluation
In User Study 2.0, the focus was on the switch between manual and autonomous driving. The study was divided into three phases: the comprehensibility of eight icons, the evaluation of four design variants for the handover process in video sequences, and detailed interviews. The data collected is crucial for optimizing the avatar and VR application and improving usability in autonomous driving.
CONCLUSION
Groundbreaking
Advances
In conclusion, the findings from the user studies were successfully integrated into optimizing the level transition and the user interface, reaching a new level of quality. Our commitment remains focused on ensuring a comfortable and safe driving experience for all passengers. At a final event at the University of Stuttgart, the consortium partners presented all the groundbreaking results of the KARLI user study. Special thanks go to the Federal Ministry for Economic Affairs and Climate Action and all consortium partners for their valuable support in this groundbreaking research project.
Process
Scenarios
Concepts
Optimization
To meet the demands of different automation levels, the teams analyzed usage scenarios covering both compliant and non-compliant driving behavior. In an internal workshop at studiokurbos, initial concept ideas were developed based on the functional sketch of the Mode Match Creator by Dr. Peter Rössger. "How might we..." questions were formulated to provide optimal support in relevant situations and address the specific challenges of each automation level.
More information about the KARLI research project can be found here.
Projects
