(22-000003193) Identifying and Predicting Inflection Points in Human-Agent Teams Using Relational Event Modeling
Abstract
A key aspect of team success is the ability to navigate changing circumstances. Human-agent teams (HAT) offer the potential for enhanced flexibility and robustness by leveraging complementary roles of human and autonomous agent teammates (AA). However, optimizingteam collaboration and coordination requires the ability for teammates to adapt to events, for example, by changing roles, strategies, or even goals. To aid this process, we aim to identify and predict inflection points, or changes in a team#s work or interactionstate, to improve teams# ability to effectively transition between states. We will use a set of experiments using simple tasks addressing different components of H-A teamwork, to induce triggering events and subsequently detect and predict the resulting inflection points across a range of scenarios. We will: 1. employ relational event modeling (REM) to associate a range of individual and group behavioral markers (physiological, neural, and verbal) with inflection points during HAT activities; 2. use machine learning to predict inflection points and 3. create a dynamic REM model, with the ultimate goal of improving teams# skills in navigating these transitions.Approved for public release
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 29, 2023
- Source ID
- N000142312420
Entities
People
- Andrea Won
Organizations
- Cornell University
- Office of Naval Research
- United States Navy