In the rapidly developing world of automated driving, our research focuses on two key challenges: ensuring safe monitoring tasks while delivering a personalized user experience.. The key to both challenges lies in the precise detection of the user’s state. In the SALSA project (Smart, Adaptive and Learnable Systems for All), we are focusing on the development of innovative technologies that use artificial intelligence to take user state detection, sensor fusion and adaptive human-machine interface to a new level.
Our vision and goals
Our main goal is to develop an advanced system for detecting the current user state. This technology should serve as a central basis for decision-making for the entire vehicle system and for human-machine interaction. We take various aspects into account:
- Mental and health status of the user
- Likelihood of falling asleep
- Physical state during sleeping or relaxation
- Detection of readiness to drive
This comprehensive state detection enables personalized adaptation of the interior, including displays, controls and seat positions.
Approaches
To achieve our ambitious goals, we are pursuing several innovative approaches:
- Expanding driver state detection: We are integrating new parameters such as attention, stress and overload into existing technology. Using advanced machine learning, we train AI systems to precisely assess individual readiness to drive.
- Advanced sensor fusion: By combining interior radar and video-based methods, we achieve an even more accurate estimation of the user’s state.
- Holistic view of the sleep process: We analyze not just sleep itself, but also the pre-sleep and post-sleep phases to gain a complete picture.
- Comprehensive user studies: Theoretical and practical research into relaxation and sleeping postures provides the basis for optimized takeover procedures.
- Analysis of takeover quality: We are investigating how different relaxation and sleeping postures influence the takeover process and its quality.
Our Contribution to the Overall SALSA Goal
Detecting the user’s state is a central building block for the SALSA project and contributes to the overall goal in a variety of ways:
- Increased safety: By precisely knowing the current user state, the vehicle can adapt its behavior and significantly increase safety, especially in takeover situations.
- Personalized interior adaptation: Based on the detected state, the vehicle interior can be adapted to the current user’s needs in a way that is appropriate to the situation and tailored to the individual.
- Intelligent knowledge transfer: New vehicle functions can be optimally communicated to the user depending on their current use and receptivity.
- Automated personalization: Driving behavior can be automatically adapted to the user’s state, which increases both comfort and safety.
- Holistic sleep concept: Through in-depth research on sleep in automated vehicles, we establish key requirements for both the interior design and overall vehicle system.
SALSA: the ideal framework
The SALSA project provides the perfect context for our research on user state recognition. It enables us to take a holistic view of and optimize various aspects of automated driving:
- Exploiting synergies: By taking user state into account, we can create synergies in related areas such as sleep concepts, knowledge transfer and automation acceptance.
- Interdisciplinary collaboration: SALSA brings together experts from different fields, which leads to innovative solutions.
- Practical research: Our close connection to the automotive industry enables us to incorporate our research results directly into the development of future vehicles.
- Future-oriented: With our focus on user state detection, we address key challenges of automated driving and make an important contribution to the mobility of the future.
Through our work in the SALSA project, we are helping to ensure that automated vehicles are not only technologically advanced, but also safe, comfortable, and individually adaptable. We are convinced that the precise recognition and consideration of the user’s state is the key to a new era of automated driving – an era in which vehicles adapt perfectly to the needs of their occupants, thus enabling a completely new driving experience.
