Research on Human-machine Dynamic Trust Based on Alarm Sequence

Abstract

With the extensive application of automation equipment, the issue of trust in automation has become a research hotspot in the field of human-machine interaction. Too high or too low trust levels may affect the performance and security of human-machine interaction. False alarm is considered to be an important factor affecting operator's automatic trust. Many experimental results show that the false alarm rate is significantly correlated with trust level. This paper studies the relationship between the alarm sequence (composed of a series of correct alarm and false alarm events) and trust level. Based on the cockpit simulation experiment, the dynamic change of the trust level with the warning event is obtained. The experimental results show significant recency and primacy effects.