A communication paradigm for human-robot interaction during robot failure scenarios

Abstract When autonomous robot systems experience failures, communication about the failure to both the people responsible for the robot and the people who happen to be nearby is critically important. New robot users, as well as bystanders, might not be familiar enough with a system to tell whether a robot is working properly and experienced robot operators might not notice signs of trouble due to being out of the loop. Thus, an important feature for robots will be the ability to communicate failure to humans when failures occur. In this chapter, we describe a study conducted with a smartphone-based feedback system and we designed to explore push and pull forms of communication. We found the communication methods improved participants’ understanding of robots’ state, increased their confidence in interacting with the robots, and allowed them to remotely monitor and control the robots. In keeping with the theme of this book, shared context between humans and machines improves performance.