Assessing Variable Levels of Delegated Control – A Novel Measure of Trust

Abstract

Autonomous cars are set to drastically change the driving environment. The promise of a safer and more efficient driving experience has led to a significant rise in research surrounding human interaction with autonomous systems, however we must investigate ways to effectively integrate these systems and develop the partnership between human and autonomous system. In particular, understanding the nature of human-automation trust will ensure safe and efficient integration of these systems, and therefore investing in new measures of trust is key to the development of the human-automation partnership. This paper discusses findings of an experiment that examines the nature of human-automation interaction and the neural correlates associated with trust. Participants were asked to interact with unmanned vehicle control stations of varying levels of control and integrity, whilst prefrontal cortical activity was monitored using functional Near Infrared spectroscopy. The findings of this study suggest that the anterior prefrontal cortex (aPFC) is associated with uncertainty of the decision-making abilities of an autonomous system, whilst the ventrolateral prefrontal cortex (vlPFC) has been implicated in the development of distrust as a result of poor decision making. The findings present a new opportunity to develop a reliable measure of human-automation trust that could inform future system design and facilitate a safer and more effective human automation partnership.