While a simple system modeling emotions can make a robot show believable emotions, it can also exhibit socially questionable or unacceptable emotions, like greeting a visitor while being visibly angry. To remedy this problem, the emotions should automatically be regulated with regards to the current situation, in order to produce more socially appropriate behaviors. To appear more natural and further increase the believability of the emotions, the robot should also exhibit behavior to cope with negative emotions.

The goal for this bachelor thesis is to extend the existing emotion system from the RicBot robots in the HelloRic project with a simple rule-based system for response-focused emotion regulation. This will allow the robots to regulate which emotion is appropriate for the given situation and be able to take actions to cope with negative or extreme emotions.

Additionally, the implemented system is going to be validated using a user study, in which the participants will have to interact with the robot with and without the added regulation mechanism. The results can then be analyzed to evaluate how the added emotion regulation changes the perceived believability and appropriateness of the shown emotions.