In current vehicles, one is increasingly confronted with new or changing systems. The number and complexity of assistance and automation systems, as well as other functions such as infotainment, will continue to increase steadily. On the one hand, this presents users with challenges when purchasing a new vehicle, and on the other hand, when taking advantage of modern mobility concepts that require users to familiarize themselves with and adjust to the operation of the respective system each time they use it. Functions often differ even in different vehicles from the same manufacturer, and there is a lot of change in relatively short periods of time. For some time now, it has been possible to update vehicle systems via the internet. These updates are becoming more and more common and thus change the familiar operating method and range of functions. It is becoming increasingly difficult to convey sufficient mental models of complex functions using only a user manual. To use automated driving safely, knowledge is needed not only about how to operate the system itself, but also about its range of functions, requirements and system limits.
In the SALSA project (Smart, Adaptive and Learnable Systems for All), several partners from industry and research are therefore working on learning concepts that meet the increased requirements of automated driving.
Our goal: user-friendly access to automated mobility
Even today, dealing with the various driver assistance systems in a new car can be exhausting and overwhelming. The complexity of these systems will also continue to increase on the road to automated driving. At the same time, safety risks can arise if the limits of the automation functions are not taken into account. Accidents have already occurred because drivers did not fulfill their responsibility to monitor the system.
When developing and evaluating various in-vehicle knowledge transfer concepts, several objectives are therefore pursued:
- Building and maintaining correct mental models of the operation, capabilities and limitations of the driving automation used
- Avoiding incorrect use of the driving automation to increase safety
- Building trust in automation without under- or overestimating functionality
- Providing the best possible user experience to increase acceptance
- User-friendly and comprehensible communication
- Consideration of the current situation and prior knowledge in order to convey information as needed
- Detection of the driver’s state of mind in order to convey learning content when people are receptive
How do we want to achieve our goal?
In order to achieve the best possible result, we follow a systematic approach in the following steps:
Recording the current state of science
As a solid basis for concept development, comprehensive literature analyses are carried out and the following questions, among others, are clarified: Which aspects of in-vehicle learning have already been researched? Which knowledge acquisition models from psychology can be successfully applied to driving automation? What are the prerequisites for learning quickly and effectively? How can learning success be verified?
Gathering user needs
In survey studies and interviews, we investigate how users have learned in the vehicle so far, how they would like to learn, and what requirements they have of a learning system.
Concept development
The results from the literature analyses and the user surveys are evaluated. On this basis, the most promising solutions are developed as concepts and implemented in the driving simulator or real vehicles. In addition, aspects are identified that require further scientific investigation.
Scientific evaluation
In the context of several studies, test drivers experience the new concepts in different variants. Based on the results of the learning success and the evaluation by the participants, the concepts are further developed and evaluated again.
How we contribute to the overall SALSA goal
During the research project, a holistic approach is being pursued to meet the challenges posed by the increasing number of automated vehicles on the road by means of smart, adaptive and learnable systems for all (SALSA). In the area of knowledge transfer, the interaction between users and their automated vehicles is the main focus. By focusing on learnability and the communication of all necessary knowledge, deficits in communication, excessive demands on users and a loss of acceptance should be counteracted in a targeted manner. An increase in road safety and the user experience are central sub-goals for all partners involved. The interdisciplinary collaboration of experts from industry, design and various scientific fields favors synergy effects and innovation.
