Understanding Expectations of a Robot ' s Identity Through Multi-User Interactions

We conducted two experiments looking at how to read user expectations of a robot's identity within multi-user environments. Multi-user environments are unpredictable and fast-paced, which can become a challenge for roboticists to interpret. However, they also present a rich landscape of data, and we propose methodologies to retrieve user reactions to the robot through sensor data. We also emphasize the necessity of using the results from these methodologies to define a robot's identity based on user expectations. In Experiment 1, we found that sensor data taken from a handshake with the robot can be used to find differences in the views of different demographic groups towards interaction with our robot. In Experiment 2, we expand the subject pool and reaffirm the usefulness of sensor data in multiuser environments, while also using questionnaire data to create an identity for our robot. I.2.9 [Robotics]: Operator interfaces – methods to research user expectations and intuitive interaction, multi-user environments, robot identity, biofeedback