COGNITIVE TRUST-BASED TASK ASSIGNMENT IN HUMAN-MACHINE TEAMING
Abstract
As agents enabled by artificial intelligence (AI) are becoming ubiquitous worldwide, human-machine teaming (HMT) is gaining increased attention. Although considerable progress has been made in balancing the distribution of control authority between humans (H) and machines (M), there continues to be a limited understanding of the effects of differences in covert (internal) states between humans and machines on this control relationship and system performance. Covert states are usually associated at different levels of uncertainty stemming from the fact that the human mind, machine sensing, and interactive environments are nonstationary in nature. However, current tools for monitoring the relationship between system performance and covert states lack systematic, efficient, and interpretable mechanisms. These problems result in a lack of reliable decision-making in HMT. This project aims to (1) develop a framework to extract and represent the covert state of humans and machines; and (2) effectively translate extracted information into trust metrics for cognitive trust-based task assignment in HMT for effective decision-making.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 04, 2023
- Source ID
- FA23862210042
Entities
People
- Tzyy-Ping Jung
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of California, San Diego