Abstract
Trust in an automated system is characterized by the expectation that it will support a person in a situation characterized by uncertainty and vulnerability. It is therefore important to know in which situation one should rely on an intelligent function and when not. If the reliability of the intelligent function is underestimated or overestimated, i.e. if it is not "calibrated" well enough, it can lead to distrust or overtrust. If these phenomena occur frequently, there can be a negative impact on the long-term acceptance of intelligent applications based on advanced AI and knowledge engineering approaches. Different elements and techniques can support the calibration of trust, but their effectiveness has so far not been systematically investigated across application domains. This paper provides an overview of the state of the art on the communication of reliability, uncertainty, awareness and intent, as well as of alternatives. Furthermore, it provides first directions and an outlook into exploiting these approaches for the calibration of trust in the application area of automated driving.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the Workshop on Automation Experience across Domains co-located with the ACM Conference on Human Factors in Computing Systems (CHI 2020) |
| Number of pages | 7 |
| Publication status | Published - 2020 |
| Event | Workshop "Automation Experience across Domains" in conjunction with CHI2020 - Duration: 26 Apr 2020 → … |
Conference
| Conference | Workshop "Automation Experience across Domains" in conjunction with CHI2020 |
|---|---|
| Period | 26/04/20 → … |
Research Field
- Former Research Field - Capturing Experience
- Former Research Field - Data Science
Keywords
- Trust calibration
- trust
- reliability uncertainty
- awareness
- artificial intelligence
- predictive systems