Robot Capability and Intention in Trust-Based Decisions Across Tasks

Abstract

In this paper, we present results from a humansubject study designed to explore two facets of human mental models of robots-inferred capability and intention-and their relationship to overall trust and eventual decisions. In particular, we examine delegation situations characterized by uncertainty, and explore how inferred capability and intention are applied across different tasks. We develop an online survey where human participants decide whether to delegate control to a simulated UAV agent. Our study shows that human estimations of robot capability and intent correlate strongly with overall self-reported trust. However, overall trust is not independently sufficient to determine whether a human will decide to trust (delegate) a given task to a robot. Instead, our study reveals that estimations of robot intention, capability, and overall trust are integrated when deciding to delegate. From a broader perspective, these results suggest that calibrating overall trust alone is insufficient; to make correct decisions, humans need (and use) multi-faceted mental models when collaborating with robots across multiple contexts.

Multi-task trust transfer for human–robot interaction

“…Arguably, trust is most crucial in new and uncertain situations whereby both robot capability and intention can influence outcomes. In very recent work (Xie et al, 2019), we have begun to examine how human mental models of these factors influence decisions to trust robots.…”

“…We believe our results to remain valid when the robot’s actions affect the human participant’s goals. Our recent work (Chen et al, 2018; Xie et al, 2019) includes behavioral measures, such as operator take-overs and greater forms of risk (e.g., in the form of performance bonuses/penalties); these experiments also provide evidence for trust transfer across tasks.…”

“…We adopt the definition of trust as a psychological attitude (Castelfranchi and Falcone, 2010) and focus on trust in robot capabilities , i.e., the belief in a robot’s competence to complete a task. Capability is a primary factor in determining overall trust in robots (Muir, 1994; Xie et al, 2019), and this work investigates how trust in robot capabilities varies and transfers across a range of tasks.…”

Trust in Robots: Challenges and Opportunities

“…Similarly, robots that came across as transparent, either by providing explanations for its actions [48,49] or simply by providing more information about its actions [50,51], were judged as more trustworthy. At a more cognitive level, one study suggests that robots that took risky actions in an uncertain environment were viewed with distrust [52], although this depends on the individual user's risk appetite [53]. In other work, robots that expressed vulnerability [54] or emotion [55] through natural language were trusted more.…”

“…Some recent works have begun to explore this area. For example, [53] examined two different aspects of trust-trust in a robot's intention and trust in a robot's capability-and demonstrated that these two factors interact to give rise to reliance on the robot. Similarly, [63] demonstrated that trust can be modeled as a latent function (i.e., an infinite-dimensional model of trust) to incorporate different contexts and showed that this approach could capture how trust transfers across different task environments.…”