Smart Automotive Mobility

The goal for the distant future is fully autonomous vehicles, which can handle every possible situation and change the role of the driver to that of a passenger. On the way to that goal, drivers of highly automated cars will still have to take over the control of the vehicle at the boundaries of the operational domain of the automation. However, several studies show that the driver’s ability to take over depends on his current status and activities. As for highly automated cars more activities beyond the actual driving task, like reading and writing emails, will be allowed, the driver might need of assistance from the vehicle automation during the control transition phase. Ironically, the fact that higher levels of automation result in decreased driver responsibilities makes considering the driver in the automation algorithms even more essential. Therefore, this chapter discusses how a human-centred control transition can be designed. Another aspect we also want to discuss is how to adapt the driving of highly automated cars to the passengers in the normal full autonomous driving scenario. Hence, we propose a novel concept of a human-centred highly automated car. First, we present an approach for monitoring and interpreting the driver in the interior and a systematic design of a cooperative control transfer. We further point out the crucial aspects of the implementation of such kind of adaption concepts and present results from various driving studies.

Automated driving is expected to bring several benefits such as improved traffic safety, reduced congestions and emissions, social inclusion, enhanced accessibility and higher driving comfort. A central human–machine interaction issue addresses the question of how automated vehicles should drive to ensure comfort and a positive driving experience. The project KomfoPilot aimed at investigating, assessing and enhancing comfort during automated driving by two driving simulator studies and a test track study. Sensors such as wearable devices, eye tracking, face tracking and motion tracking allowed for the integration of driver state data with information about the vehicle and the surroundings. Various driving styles as well as display solutions were evaluated for reducing discomfort. In addition, privacy issues were continuously monitored for all aspects over the project lifetime. Section 2.1 gives an overview on the background and aims of the project, definitions of central concepts and an overall summary of key results. Methodological details on the experimental design, participants, assessment of discomfort, sensors and questionnaires of the three studies are presented in Sect. 2.2. A first research objective was to find factors that affect comfort on a rather general level, such as driving situations and driving style parameters (e.g. speed, longitudinal/lateral distance, driving style familiarity). These results are presented in Sects. 2.3 and 2.4. A second objective was the development of algorithmic approaches for real-time discomfort detection based on sensor data. Results on physiological discomfort indicators are presented in Sect. 2.5, whereas Sect. 2.6 gives an overview of the work on algorithms. Privacy and liability aspects are discussed in Sect. 2.7.

Imagine you are at the beginning of a journey from Stuttgart to Munich. It will take you almost three hours because of heavy traffic. It is quite warm outside; you have just had lunch and feel a bit tired. This is sure to be a long, exhausting, and boring trip. The good news is that your car can drive automatically and that you have KoFFI (in German: “Kooperative Fahrer-Fahrzeug-Interaktion”) on board—the new intelligent driver assistance system for collaborative driving in both manual and automated driving modes. In this chapter, we describe how KoFFI supports you in typical traffic situations during that drive. On the one hand, there is the so-called guardian angel function, which helps you to survive critical traffic situations but also offers some convenient features during manual driving. On the other hand, you will learn how KoFFI can assist the driver at system boundaries and vice versa in various cooperative driving scenarios. In addition, we explain how to apply ethics-by-design during system development and how to take care of your personal data required for automated driving (e.g., driver monitoring video streams or data needed for personalization). KoFFI communicates with the driver via its innovative speech dialogue system, which can even distinguish between priorities and a user-centered human-machine interface. The results of and lessons learned from several user tests show that the cooperative assistant KoFFI is able to ensure a convenient, pleasant, and safe drive in either manual or automated driving mode.

The KoFFI guidelines have been formulated during the project duration. Even though in the meantime, many guidelines have been published, we consider it as relevant, that everybody reflects about their doings and about their own ethical principles. As a means to achieve this, we offer a questionnaire for ethically reviewing a project. In a dialogue with the project partners, we have defined four values which we consider as central for the development of highly automated driving in general and the KoFFI project in particular. These values are privacy, autonomy, trust (in technology) and transparency. With the questionnaire, we suggest a tool to perform a self-check of ethical issues that might arise while doing research on highly automated driving. Slightly adapted, the questionnaire can also be used for the development of other technologies. It can be used during quality management procedures and should be reused during the ongoing of the project. Having completed the questionnaire researchers, developers and the different actors involved will probably agree that taking into account ethics can represent and advantage for the respective company.

Automated driving is dramatically changing the way we perceive and control cars and lorries. In partially automated driving, the driver is still responsible for the vehicle control, but is strongly supported. In highly automated driving, the driver can give control and responsibility to the automation for a certain time and can get control back, e.g. when the automation encounters limits. However, an intuitive way to interact with these automated modes is largely missing. Vorreiter is addressing this by using the inspiration of a rider and a horse to provide intuitive steering gestures on the steering wheel or an alternative device, which initiate manoeuvres to be executed by the automation and to be supervised, influenced or interrupted by the driver. The gestures are built up in a design-for-all which helps all drivers, including beginners and drivers with disabilities. A consortium of RWTH Aachen University, Fraunhofer IAO, University of Stuttgart, Valeo and Hochschule für Wirtschaft und Recht Berlin defined a concept, built up prototypes and investigated its impact in workshops and simulator experiments.

From the perspective of traffic safety, interactions between cars, bicyclists and pedestrians that lead to crashes comprise the largest part of all crashes. One major cause is missing or inadequate communication. This results in inadequate understanding and anticipation of what the other is going to do. Additionally, communication could further cooperation which leads to positive emotions and a better traffic climate. Within the project KOLA (Kooperativer Laserscheinwerfer; cooperative laser beam), we first examined needs in traffic which lead to positive emotions (positive arousal, being thrilled and feeling close to others) when fulfilled, and negative, if denied (being irritated, annoyed, horrified and feeling hostility). With regard to communication messages, explaining one’s intentions and actions and thanking others was identified as the most relevant topics. Prototypical messages were designed and evaluated in two driving simulator studies using light which was projected from a car as the medium of the messages. We found that expressing one’s intention in a specific situation furthers understanding and prosocial behavior of others, which leads to positive emotions on both sides. Explaining one’s unusual behavior works similarly to increase understanding and further positive emotions but seems to require specific information about the causes of the behavior. Enabling drivers to thank others encourages prosocial behavior and leads to positive emotions. Thus, light-based communication is an excellent way to further communication, cooperation and positive emotions in traffic. The studies presented provide a sound starting point for further research and development.