Neural Correlates of Trust in Automation: Considerations and Generalizability Between Technology Domains

Abstract

Investigations into physiological or neurological correlates of trust has increased in popularity due to the need for a continuous measure of trust, including for trust-sensitive or adaptive systems, measurements of trustworthiness or pain points of technology, or for human-in-the-loop cyber intrusion detection. Understanding the limitations and generalizability of the physiological responses between technology domains is important as the usefulness and relevance of results is impacted by fundamental characteristics of the technology domains, corresponding use cases, and socially acceptable behaviors of the technologies. While investigations into the neural correlates of trust in automation has grown in popularity, there is limited understanding of the neural correlates of trust, where the vast majority of current investigations are in cyber or decision aid technologies. Thus, the relevance of these correlates as a deployable measure for other domains and the robustness of the measures to varying use cases is unknown. As such, this manuscript discusses the current-state-of-knowledge in trust perceptions, factors that influence trust, and corresponding neural correlates of trust as generalizable between domains.

Trust in Shared-Space Collaborative Robots: Shedding Light on the Human Brain

“…Thus, manipulation of robot reliability is the treatment to both survey responses and neural activity, which were both found to be sensitive to the construct of trust. Identifying neural activity in response to changes in subjective perception alone did not result in any identified region, likely due to the variance in human response when answering the questions on the TRUST survey or validity concerns with surveys, where trust experience may not be fully capturable by the provided questions (Hopko, Mehta et al, 2021). Because subjective and neural response provide different types of information and are both symptoms of trust manipulation, identifying trust correlates based on the subjective responses risks losing the sensitivity of neural responses to aspects of trust that surveys cannot capture.…”

“…To the authors' knowledge, no prior works identify neural correlates of trust in shared space HRC. Rather, past work identifying neural correlates of trust focused on cyber aid systems including sensor feeds, alerts, and recommendation systems with simple action and failure spaces (Hopko & Mehta, 2021). Thus, trust is often manipulated by the correct and incorrect response rate of these tools, easily identifiable by the user.…”

“…Because subjective and neural response provide different types of information and are both symptoms of trust manipulation, identifying trust correlates based on the subjective responses risks losing the sensitivity of neural responses to aspects of trust that surveys cannot capture. It can be difficult to quantify subconscious processes, such as changes in cognitive processing associated with trust (Mayo, 2015) in survey responses, whereas explanation of trust-related response to robot behaviors can be explained through the use of physiological monitoring (Hopko & Mehta, 2021).…”

“…It is important to consider that this study was conducted on a novice student population which has a higher education level than traditional manufacturing workforce (USA Bureau of Labor Statistics, 2020), although technicians and operators of cobots may require higher educations in order to have fluent collaboration. As this study was conducted on persons with limited-to-no experience with collaborative robotics, understanding the impact of experience, a known factor influencing trust attitudes and trust dynamics (Hopko, Mehta et al, 2021), similar investigations utilizing a workforce with more experience is warranted. Specifically, this experience factor may be relevant as experience may impact the cognition required to identify nominal verses off-nominal robot behaviors and cognition required to adapt, thus changing the magnitude of perceptual and neural response or the specific adaptation strategies utilized.…”

“…Like many other subjective states, the state-of-the-art method to capture trust perceptions is through surveying or interviewing end-users (Hopko et al, 2022). Beyond the benefits of subjective questionnaires alone, neural responses provide rich spatiotemporal mechanistic understanding of trust influencers to provide unique insights into subsequent human behavior (Hopko & Mehta, 2021). The utilization of neural activity may help reveal why changes in trust occur through identification of activation/suppression of functionally varied neurons and shifts in integration and efficiency of neural connectivity.…”