User Evaluation of Adaptive Attention-attracting Behavior of robot

In the future, it is expected that robots will act as our companions by taking some load off our minds and making our lives more efficient and easy. One of their functions would be to provide assistance to people such as giving reminders for medicines, appointments, weather forecast, etc. To initiate communication, these robots would first need to attract a person’s attention by performing some behavior. This study assumes that different people would prefer the robot to perform different behaviors to attract their attention. To adapt to a particular person’s preferences, the robot must learn from the reactions to its behaviors in different situations. The aim of this study is to test whether the learnt behaviors-situations combinations of the robot makes the task of attracting attention more time-efficient and further the person’s impression of the robot more positive after learning and adapting. The reported results suggest no significant effects in terms of reaction times to the learnt behaviors and no significant change in impression of the robot when it learns. However, the study does give some insights of used experimental methods and the workings of the used learning algorithm.