An Investigation of Human-Machine Team Dynamics
Abstract
Smart technology plays an increasingly large role in society today, and as a result humans are interfacing with artificial intelligence (AI) more frequently and in more complex ways. While there has been a significant focus on how this technology can change industries and alter the jobs of the future, there has been less emphasis on how AI can be effectively integrated into a team. Because teams are the fundamental building blocks of organizations, including the government, military, and industry, understanding how these human-machine teams function is of broad importance. The proposed research is an investigation of two interrelated elements of human- machine team functioning: team processes, team emergent states, and their joint dynamics. Human-machine team processes include all of the discrete actions and interactions which transpire during the accomplishment of team tasks. Human-machine team emergent states are psychological constructs which develop over time as a consequence of team processes, endogenous team attributes, and exogenous environmental factors. Together, these elements describe a team cognitively, behaviorally, and affectively. The objective of this research is to better understand how the inclusion of an intelligent machine agent changes the processes and states of a human team, and what implications those changes have for performance. In order to study these phenomena, the relational event network modeling framework will be applied. A relational event model (REM) is a statistical model designed to detect patterns in sequences of discrete events. Recent advances have demonstrated the efficacy of this approach for describing team processes, connecting events to emergent states, and explaining team performance. To conduct relational event analysis, granular data Ð e.g., a log of team actions and interactions Ð coupled with survey data is needed. As such, data for this project will be obtained from a combination of laboratory and field experiments. The lab setting will provide a controlled setting to observe how an AI teammate will impact human participates while they complete a variety of problem-solving tasks. The field setting will be an opportunity to deploy an AI collaborator into real-world scenarios. The proposed research constitutes a significant contribution to the scientific state-of-the-art in the field of human-machine teaming. A deeper understanding of how humans interact with autonomous technology would both help existing organizations shape their technology strategy, as well as suggest new directions for future research. Additionally, there are several potential implications of this research which would benefit the Army and Department of Defense. While humans are still critical components of any military mission, smart technologies are likely to play an increasing role. Autonomous agents may help to improve the efficiency of teams, increase the likelihood of successful outcomes, and enhance the safety and well-being of humans in the field. Finally, the proposed research will make methodological contributions through the extension of existing models, the development of new analytical tools, and the production of statistical software for future use.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jul 24, 2019
- Source ID
- W911NF1910427
Entities
People
- Aaron Schecter
Organizations
- Army Contracting Command
- The University of Georgia
- United States Army