Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enabling Team of Teams: A Trust Inference and Propagation (TIP) Model in Multi-Human Multi-Robot Teams

Yaohui Guo, X. Jessie Yang, Cong Shi

TL;DR

The TIP framework addresses trust in fluid, multi-agent teams by modeling trust as a Beta-distributed variable updated from both direct robot interactions and indirect propagation via teammates. Theoretical results guarantee convergence to a unique equilibrium under repeated interactions, and parameter inference is performed via maximum likelihood with gradient descent. Empirical evaluation with 30 participants across 15 teams shows TIP captures trust dynamics more accurately than baselines and demonstrates faster within-team convergence and reduced trust deviation due to trust propagation. The approach enables trust estimation in networks with many humans and robots, offering a principled method to fuse direct and indirect experiences for robust HRI trust management.

Abstract

Trust has been identified as a central factor for effective human-robot teaming. Existing literature on trust modeling predominantly focuses on dyadic human-autonomy teams where one human agent interacts with one robot. There is little, if not no, research on trust modeling in teams consisting of multiple human agents and multiple robotic agents. To fill this research gap, we present the trust inference and propagation (TIP) model for trust modeling in multi-human multi-robot teams. In a multi-human multi-robot team, we postulate that there exist two types of experiences that a human agent has with a robot: direct and indirect experiences. The TIP model presents a novel mathematical framework that explicitly accounts for both types of experiences. To evaluate the model, we conducted a human-subject experiment with 15 pairs of participants (${N=30}$). Each pair performed a search and detection task with two drones. Results show that our TIP model successfully captured the underlying trust dynamics and significantly outperformed a baseline model. To the best of our knowledge, the TIP model is the first mathematical framework for computational trust modeling in multi-human multi-robot teams.

Enabling Team of Teams: A Trust Inference and Propagation (TIP) Model in Multi-Human Multi-Robot Teams

TL;DR

The TIP framework addresses trust in fluid, multi-agent teams by modeling trust as a Beta-distributed variable updated from both direct robot interactions and indirect propagation via teammates. Theoretical results guarantee convergence to a unique equilibrium under repeated interactions, and parameter inference is performed via maximum likelihood with gradient descent. Empirical evaluation with 30 participants across 15 teams shows TIP captures trust dynamics more accurately than baselines and demonstrates faster within-team convergence and reduced trust deviation due to trust propagation. The approach enables trust estimation in networks with many humans and robots, offering a principled method to fuse direct and indirect experiences for robust HRI trust management.

Abstract

Trust has been identified as a central factor for effective human-robot teaming. Existing literature on trust modeling predominantly focuses on dyadic human-autonomy teams where one human agent interacts with one robot. There is little, if not no, research on trust modeling in teams consisting of multiple human agents and multiple robotic agents. To fill this research gap, we present the trust inference and propagation (TIP) model for trust modeling in multi-human multi-robot teams. In a multi-human multi-robot team, we postulate that there exist two types of experiences that a human agent has with a robot: direct and indirect experiences. The TIP model presents a novel mathematical framework that explicitly accounts for both types of experiences. To evaluate the model, we conducted a human-subject experiment with 15 pairs of participants (). Each pair performed a search and detection task with two drones. Results show that our TIP model successfully captured the underlying trust dynamics and significantly outperformed a baseline model. To the best of our knowledge, the TIP model is the first mathematical framework for computational trust modeling in multi-human multi-robot teams.
Paper Structure (22 sections, 3 theorems, 46 equations, 12 figures)

This paper contains 22 sections, 3 theorems, 46 equations, 12 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Trust Modeling in Dyadic Human-robot Interaction
  4. Reputation/Credential Management
  5. Mathematical Model
  6. Assumptions
  7. Proposed Model
  8. Asymptotic Behavior under Repeated Interactions
  9. Parameter Inference
  10. Trust Estimation
  11. Human Subject Study
  12. Participants
  13. Experimental Task and Design
  14. Experimental Procedure
  15. Results and Discussion
  16. ...and 7 more sections

Key Result

Theorem 1

When $m >0$ and $n\geqslant 0$, $t_{k}^{x,A}$ and $t_{k}^{y,A}$ converge in probability (i.p.) respectively, i.e., there exists $t^{x}$ and $t^{y}$ such that, for any $\epsilon >0$,

Figures (12)

  • Figure 1: Four agents can form sub-teams. In part 1, human $x$ and robot $A$ form a dyad, and human $y$ and robot $B$ form a dyad. In part 2, two dyads merge. In part 3, human $x$ and robot $B$ form a dyad, and human $y$ and robot $A$ form a dyad.
  • Figure 2: An arrow points from a trustor to a trustee, representing the trust $t^{\text{trustor}, \text{trustee}}$. Human $x$ updates her trust in robot $B$ based on direct experience. Even though $x$ does not have direct interaction with $A$, $x$ could still update her trust toward $A$ through a third party, $y$.
  • Figure 3: $x$ and $y$ take turns to interact with $A$.
  • Figure 4: Experimental process and task interface.
  • Figure 5: Illustration of drone assignment. Participant $x$ is randomly assigned to work with drone $A$ in session 1, with drone $B$ in session 2, and so on. The assignment is random.
  • ...and 7 more figures

Theorems & Definitions (3)

  • Theorem 1
  • Theorem 2
  • Corollary 1