Robot Feedback Design for Response Delay

Abstract

We investigated an effective human-robot interaction design to alleviate the user's negative experience caused by the response delay of a robot. The feedback regarding a robot's delayed response was designed using human-like and machine-like approaches. The first experiment was conducted in task-based interaction to explore the effects of the feedback type (human-like or machine-like feedback versus a baseline robot) and task type (a high-cognitive-demand task versus a low-cognitive-demand task) on the understandability, perceived waiting time, and service evaluation of the robot (N = 36). The robot with machine-like feedback was rated the most positive in terms of understandability, perceived waiting time, and service evaluation. The perceived waiting time mediated the effect of feedback type on service evaluation. There were significant effects between the feedback types and the task types on understandability, perceived waiting time, and service evaluation. In a low-cognitive-demand task, machine-like feedback was evaluated more positively than human-like feedback, while in a high-cognitive-demand task, there was no significant difference between human-like feedback and machine-like feedback. In addition, the second experiment was conducted in social interaction between humans and robots to explore the effect of the two feedback types on the understandability, perceived waiting time, sociability, and service evaluation of the robot (N = 36). The robots providing feedback were evaluated more positively than the baseline robot, which did not provide any feedback, and sociability mediated the effect of feedback type on service evaluation.