Operationally Realistic Human-Autonomy Teaming Task Simulation to Study Multi-Dimensional Trust

Autonomous systems are increasingly integrated into performing critical tasks, including in the space domain. Humans need to appropriately trust their autonomous teammate if the team is to perform effectively. Historically, many trust studies have used obtrusive surveys to measure trust as a one-dimensional, static construct when it is actually dynamic and multi-dimensional in nature. Furthermore, some previous work uses trusting tasks that lack operational validity. This paper presents an operationally-realistic task to study dynamic, multiple dimensions of operator trust. It enables collecting and time-synchronizing biosignals and embedded measures with less than millisecond precision. Our results show that this task simulation is capable of collecting unobtrusive, multidimensional trust-relevant signals with the end goal of improving human-autonomy teaming performance.