A Literature Review of Trust Repair in HRI

Abstract

Trust is vital for effective human-robot teams. Trust is unstable, however, and it changes over time, with decreases in trust occurring when robots make mistakes. In such cases, certain strategies identified in the human-human literature can be deployed to repair trust, including apologies, denials, explanations, and promises. Whether these strategies work in the human-robot domain, however, remains largely unknown. This is primarily because of the fragmented and dispersed state of the current literature on trust repair in HRI. As a result, this paper brings together studies on trust repair in HRI and presents a more cohesive view of when apologies, denials, explanations, and promises have been seen to repair trust. In doing so, this paper also highlights possible gaps and proposes future work. This contributes to the literature in several ways but primarily provides a starting point for future research and recommendations for studies seeking to determine how trust can be repaired in HRI.

Having the Right Attitude: How Attitude Impacts Trust Repair in Human—Robot Interaction

“…Trust repairs are actions taken by a trustee to restore trust after a perceived or actual violation of trust [26,27,20]. These actions can take various forms but are often realized through short-term verbal repair strategies, such as apologies, denials, explanations, and promises [19,28,22]. Each of these trust repair strategies can be linked to different dimensions of trustworthiness and are supported by distinct overarching theoretical frameworks [13,24,21].…”

“…Apologies are expressions of regret or remorse for a perceived transgression. They are the most widely used trust repair strategy across the literature and are believed to repair trust through encouraging forgiveness [21,22]. Denials are attempts to redirect blame or reject culpability for a trust violation [29], aiming to establish the complete innocence of the trustee by shifting blame onto another entity [19].…”

“…Trust repair strategies can be defined as planned approaches or sets of actions designed to rebuild trust after it has been damaged or violated [20,21]. While researchers in human-robot interaction (HRI) have started investigating the effectiveness of these strategies when employed by robots, a recent literature review revealed mixed results regarding their efficacy [22]. Moreover, a comprehensive theoretical framework for trust repair in HRI is yet to be established, and the specific mechanisms underlying different repair techniques remain largely unexplored.…”

“…Recent studies have examined the efficacy of trust repair strategies when deployed by robots. Findings, however, have been mixed, and there is a lack of a comprehensive theoretical framework for trust repair in HRI [22]. As a result, designers and developers still lack the knowledge needed to build robots that can effectively and consistently restore trust after violations occur.…”

A Theoretical Model to Explain Mixed Effects of Trust Repair Strategies in Autonomous Systems

“…However, only fairly recently has it been given research attention (Robinette et al, 2015) and a framework for its implementation (de Visser et al, 2018). Since then, there has been a surge of studies that have examined the efficacy of different trust repair strategies in a wide variety of domains (for a partial review, see Esterwood & Robert, We are in total agreement with Esterwood and Robert (2022) that the lack of a strong theoretical foundation in the trust repair literature is a stumbling block because it limits our ability to explain current disparate findings, generalise results to other situations, and generate new research hypotheses. While we think that it may be premature to carry out a formal metaanalysis of general trust repair effects (partly due to the number of studies, but also the wide variety of trust repair strategies studied within the available studies), it may still be useful to look for trends in the landscape of findings to try to identify inconsistencies and discern patterns in the wider trust repair literature (as Esterwood & Roberts have done within HRI).…”