Emotion detection in the context of driving

Kurzfassung

With varying levels of automation in future vehicles, the role of the human user will change from an active driver to a part-time controller or mere passenger even within one journey. In order to guarantee safety and acceptance of automated vehicles, human-machine interfaces and interaction strategies tailored to the current needs of the human are required, so that the users are optimally supported in all different roles. Empathic systems that, in analogy to human empathy, adapt their behavior and communication based on the assessment and interpretation of the current state and by doing so provide individual support are a promising concept to fulfil this requirement. These, however, need a reliable recognition of the user’s current cognitive and emotional state, as well as strategies to appropriately react to it. Therefore, the talk will present the results of recent driving simulator studies investigating behavioral and (neuro-) physiological indicators of cognitive and emotional user states as basis for automated user state recognition. In addition, potential strategies on how empathic vehicles could adapt to the current user state in order to offer optimal support will be discussed.