| Document
|
Author Chris Baber, Sagir Yusuf, Edmund Hunt, Sanja Milivojevic, Patrick Waterson |
| Abstract When people and robots work together, performance can be affected by perceived trust. Typically, trust is measured through self-report from people, often at the end of an experiment. We present a simple approach to estimate trust levels from data during performance. The approach not only allows us to estimate trust that humans have in the robots they work with but also trust that the robots might have in their human teammates. We propose that this could be used to identify degradation in trust and provides opportunity to repair trust before interaction is impaired |
