Our daily lives present us with a range of sensors that enables us to access a world of technology with increasing levels of comfort. We unlock our smartphone with an iris scan, use a microphone to give voice commands to control networked systems, have our steps counted by motion sensors, measure our pulse by laying a finger on our phone’s camera, and navigate through traffic with the help of GPS systems. However, how can information from sensors be integrated to optimize technical assistance systems in vehicles, for example, in order to adaptively respond to the driver’s emotional states or social interaction? This question was the subject of the three-year research project conducted by Spiegel Institut along with four other partners under the auspices of the German Federal Ministry of Education and Research.
Mannheim, February 18 – The project “Fundamentals of interaction and emotion sensitive assistance systems”, in short INEMAS, has been promoted by the German Federal Ministry of Education and Research since 2015 to improve the interaction between humans and technology. For example, this project looked at how warning systems in vehicles can respond to social interaction in the vehicle as well as the user’s emotions in order to take into account receptiveness and responsiveness at any given moment. In addition to the technical feasibility, the project studied cost-benefit aspects as well as the acceptance of such systems and the impact on user behavior.
In addition to Spiegel Institut, the Catholic University of Eichstätt-Ingolstadt, the University of Siegen, the University of Münster as well as Audi AG were part of the research network.
After the joint final meeting in Eichstätt, at the end of January 2018, the results were presented in Bonn at the networking meeting “Social- and emotion-sensitive systems for an optimized human-technology interaction – from technical tool to interactive companion (InterEmotio)”.
Influence of emotions and social interactions on driving behavior
Overall, six studies in three experimental environments were carried out in the last three years, two of which in Spiegel Institut’s own driving simulator. While users completed driving tasks in the simulation, either emotion induction by playing music or social interaction with the passenger was introduced in the form of a question-answer game. Effects of the emotional arousal were revealed; for example, the emotion of “annoyance” caused a faster braking reaction than the emotion of “joy”. As the effects strongly depended on the driving task at hand, one can assume that motivational effects play a role (e.g., more aggressive driving style when annoyed) in addition to emotions. Regarding social interaction, as expected the braking response is fast whenever the question-answer game was not being played. Surprisingly, this advantage disappears if the users are told they are being filmed.
A look into the future
Can information on social interactions and emotions be integrated effectively into driving assistance systems? The research project shows that it is now technically possible to design driver assistance systems to be adaptive depending on the current state of the driver and her individual responsiveness. This allows for user-centered research in real time.
While the effects of the emotional state and social interaction can be observed in the driving data (e.g., braking reaction), first and foremost only the interaction can be measured by the camera. This is because the user’s facial expressions show too little variance while driving in order to categorize the emotional states.
Initial reservations decrease after system contact
While the principal intention of using social-/emotion-sensitive assistance systems in an initial survey study was still very low, the participants of the subsequent simulation studies evaluated the experienced system positively.
In a focus group at Spiegel Institut at the end of the project, the advantages and disadvantages of social-/emotion-sensitive systems from a user perspective were presented and discussed with drivers. With regard to acceptance, sensors in the vehicle to detect social interaction are still more tangible for the user than recording the emotional state of the driver. The effect on distraction in terms of body posture (turning away from the driving task) and volume in the vehicle can be seen with a strong level of social interaction.
This raises a crucial question that emerged early in the project: if you ask the user, the acceptance of a data-collecting system is greater if the value added is clearly stated and the behavior of the system is transparent. But what if these can only be effectively adapted through the complex handling of the collected information, which, in turn, cannot be presented to the user in a transparent manner?
Nonetheless, a research project would be not a research project if there was any other conclusion than “...further research is needed”.
Felix Leist, M.Sc. Psychology
Project Manager International Market Research
Spiegel Institut Mannheim GmbH & Co. KG
Hermsheimer Straße 5
Director Spiegel Institut Communication
Spiegel Institut Communication GmbH & Co. KG
Hermsheimer Straße 5
(Image source: Spiegel Institut)