Chapter 13. Trust in evacuation robots

This chapter explores how to create robots that enhance user safety by promoting an appropriate amount of trust during an emergency evacuation. The results from our experiments demonstrate that people tend to overtrust robots and, during an emergency, will follow or wait by a broken robot. Human-robot trust research often assumes that people will consider the robot’s reputation and reliability to evaluate how much they should trust it. This work shows that people may swiftly trust a robot in spite of signs that the robot is not trustworthy. Future work should be devoted to better understanding why and when people will overtrust a robot and how to prevent all types of automation bias.