Bidirectional trust modelling in Human-machine teaming in dynamic environments

Abstract

Human-Machine Teaming (HMT) is a rapidly evolving domain that enhances collaboration between human intelligence and machine autonomy across various dynamic environments. Despite significant advancements, a critical gap remains in understanding the bidirectional trust dynamics necessary for effective human-machine collaboration. Existing research primarily focuses on one-directional trust, where humans must trust machines, often neglecting the reciprocal nature of trust where machines also need to trust human operators. This gap limits the ability to develop robust and adaptive systems that can operate effectively in complex, high-stakes scenarios. To address this gap, our research focuses on developing novel algorithms that model the fundamental principles of bidirectional trust within HMT. We propose a unique temporal-spatial knowledge graph framework designed to capture and analyze the evolving relationships between humans, machines, and their tasks, enabling a dynamic calibration of trust in response to changing conditions. Additionally, we are creating innovative algorithms that integrate Bayesian methods with hierarchical reinforcement learning (RL) and Video-and-Language Models (VLMs). These algorithms will not only uncover new insights into how trust dynamics influence decision-making and policy generation in complex environments but also push the boundaries of current trust modelling techniques, offering a new perspective on trust as a bidirectional construct. Through this basic research, we aim to advance the theoretical foundations of trust dynamics in HMT, providing a novel framework that can inform future studies and pave the way for the development of more reliable and adaptable AI systems. Our work contributes to the core knowledge of human-machine interactions, laying the groundwork for innovations that enhance collaboration and trust in increasingly autonomous systems.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 06, 2025
Source ID
FA23862514001

Entities

People

  • Mohammad Rastgoo

Organizations

  • Air Force Office of Scientific Research
  • Monash University
  • United States Air Force

Tags

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Systems Analysis and Design
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy
  • AI & ML - Neural Networks
  • Autonomy
  • Autonomy - Autonomous System Control
  • Autonomy - Human-Robot Interaction