How Does Presence of a Human Operator Companion Influence People's Interaction with a Robot?

Sun, Y., & Carruth, D. W. (2018). How Does Presence of a Human Operator Companion Influence People's Interaction with a Robot? Advances in Human Factors in Robots and Unmanned Systems. Orlando, FL: Springer. 136-142. DOI:10.1007/978-3-319-94346-6_13.

Abstract

In the future, humans and robots may work together for different reasons. It is important to understand how people’s behavior towards a robot changes depending on whether the robot operator is present and nearby the robot or absent and not visible to the individual. This paper focuses on a comparison of two situations: the robot working by itself and the robot working with the operator next to the robot and following the robot around. Our study examines whether the presence or absence of a human operator changes an individual’s interest in the robot or the likelihood that an individual will exhibit dishonest behavior when interacting with the robot. A pet, when accompanying their owner, can encourage other individuals to interact socially with the owner [1][2]. The combination of a human and a dog encourages social interaction. When a human is paired with a robot, it is possible that the combination of the human and the robot is more likely to capture the interest of individuals and encourage interaction. While previous research has examined interest in interacting with anthropometric robots in public settings, the authors are unaware of previous research directly investigating the effect of the presence or absence of a human operator on interaction with the robot. In most cases, even if the robot is remotely operated, the goal of the study is to examine interaction with an apparently autonomous robot. While we expected that the combination of a human and robot might encourage interaction, we also suspected that the presence of the human would encourage honesty in the interaction with the robot. In research comparing the effect of a robot observer to the effect of a human observer, the robot observer by itself did not discourage dishonest behavior [3]. In the current study, we examine whether those who choose to interact with the robot when the human is present are more honest than those who interact with the robot when the human is absent. Our experiment used a semi-anthropomorphic robot that delivered a bowl filled with candies to the student union and the main campus library at Mississippi State University. This experiment aimed to evaluate people’s behavioral change toward the robot when the robot had a human operator nearby and when the robot appeared to be acting autonomously. A camera recorded the general area around the robot and a second camera recorded interactions with the robot's bowl of candy. The area camera was used to record the total number of people passing through the area. The number of individuals that appeared interested in the robot and the number that approached and interacted with either the operator or the robot were counted. The candy camera was used to record how many candies each individual took from the bowl. Data was collected on four days in one week. On Monday and Wednesday, the experiment was performed at the same time of day at the student union. On Tuesday and Thursday, the experiment was performed at the same time of day at the main campus library. On one day, the operator was positioned near the robot and, on the other day, the operator controlled the robot remotely and was not visible to individuals in the area around the robot. The results revealed that more individuals approached the robot when the robot was alone and acting autonomously. With respect to honesty, as expected, individuals were more likely to take extra candy when the robot was alone. It is possible that rather than encouraging interaction, the presence of the human operator increased feelings of being observed and discouraged individuals from even interacting with the robot. The results have implications for likelihood of individuals interacting with robots teamed with visible human operators and for honesty of interactions with robots that are autonomous and those with human operators.