Modeling and Predicting Trust Dynamics in Human–Robot Teaming: A Bayesian Inference Approach

Abstract

Trust in automation, or more recently trust in autonomy, has received extensive research attention in the past three decades. The majority of prior literature adopted a “snapshot” view of trust and typically evaluated trust through questionnaires administered at the end of an experiment. This “snapshot” view, however, does not acknowledge that trust is a dynamic variable that can strengthen or decay over time. To fill the research gap, the present study aims to model trust dynamics when a human interacts with a robotic agent over time. The underlying premise of the study is that by interacting with a robotic agent and observing its performance over time, a rational human agent will update his/her trust in the robotic agent accordingly. Based on this premise, we develop a personalized trust prediction model and learn its parameters using Bayesian inference. Our proposed model adheres to three properties of trust dynamics characterizing human agents’ trust development process de facto and thus guarantees high model explicability and generalizability. We tested the proposed method using an existing dataset involving 39 human participants interacting with four drones in a simulated surveillance mission. The proposed method obtained a root mean square error of 0.072, significantly outperforming existing prediction methods. Moreover, we identified three distinct types of trust dynamics, the Bayesian decision maker, the oscillator, and the disbeliever, respectively. This prediction model can be used for the design of individualized and adaptive technologies.

Trust Dynamics and Verbal Assurances in Human Robot Physical Collaboration

“…The majority of trust researchers, though, resort to post-hoc questionnaires as the main strategy to estimate and model trust since it is a latent variable and it is still challenging to measure it. This, however, typically provides a measurement of trust at a single point of time ( Guo and Yang, 2020 ) usually administered at the end of an experiment.…”

“…Hence, in order to appropriately quantify trust, its dynamics must be taken into consideration. However, most of the existing methods on modeling and measuring trust rely mainly on post-hoc questionnaires at the end of an experiment which provides only a “snapshot view” of trust ( Guo and Yang, 2020 ) instead of measuring it continuously.…”

A Unified Bi-directional Model for Natural and Artificial Trust in Human–Robot Collaboration

“…Planning and decision-making frameworks usually rely on the use of probabilistic models for trust [5], [24], [25]. Xu and Dudek proposed an online probabilistic trust inference model for human-robot collaborations (OPTIMo) that uses a dynamic Bayesian network (DBN) combined with a linear Gaussian model and recursively reduces the uncertainty around the human operator's trust.…”

“…Other Bayesian models have been proposed since OPTIMo. These models include personalized trust models that apply inference over a history of robot performances, such as [25] and [24]. Mahani et al proposed a model for trust in a swarm of UAVs, establishing a baseline for human-multirobot interaction trust prediction [25].…”

“…Mahani et al proposed a model for trust in a swarm of UAVs, establishing a baseline for human-multirobot interaction trust prediction [25]. Guo and Yang [24] have improved trust prediction accuracy (as compared to Lee's ARMAV model [13] and OPTIMo [5]) by proposing a formulation that describes trust in terms of Beta probability distributions and aligns the inference processes with trust formation and evolution processes [24]. Without explicitly modeling trust, Lee et al showed that a robot that estimates and calibrates humans' intents and capabilities while making decisions can engender higher trust from humans [27].…”