Using Haptic Feedback in Human Robotic Swarms Interaction

Abstract

Robotic swarms display emergent behaviors that are robust to failure of individual robots, although they can not necessarily accomplish complex tasks with these behaviors. The research objective is to make use of their robust behaviors to accomplish complex tasks in many types of environments. For now, it is difficult to affect swarm “goals”, and therefore difficult them to direct to perform complex tasks. The extant literature on Human Swarm Interaction (HSI) focuses on demonstrating the usefulness of human operator inputs for swarms to accomplish complex tasks. The human typically gets visual feedback of the state of the swarm and influences the robots through a computer interface. This paper presents a user study investigating the effectiveness of haptic feedback in improving HSI. We use methods developed in studies using haptics in multi-robot systems (where the communication and structure is very rigid) and potential field algorithms developed for fully-autonomous swarms to determine the benefits of haptic feedback from the semi-autonomous control algorithm. In some environments, haptic feedback proved beneficial whereas in other environments haptic feedback did not improve performance over visual feedback alone. However, presence of haptic feedback did not degrade the performance under any of the experimental conditions. This supports our working hypothesis that haptic feedback is potentially useful in HSI.