The Impact of Perceived Autonomous Agents on Dynamic Team Behaviors

Mustafa Demir, Nathan J. McNeese, Nancy J. Cooke

Research output: Contribution to journalArticlepeer-review

23 Scopus citations


Adaptive complex team behaviors evolve dynamically and occur in many different environments. In this study, we examined the role of these behaviors and their relationship with team performance in the context of human-autonomy teams (HAT) and all-human teams. The HAT served as the "synthetic" condition in which two human team members were informed that the third team member was a "synthetic" agent; in the control condition, the team members were informed that the pilot was a remotely located human teammate. Following are the primary findings from this study: first, control teams demonstrated better performance than the synthetic teams; second, control teams were more active than the synthetic teams in terms of planning the task, and third, the behavioral passiveness of the synthetic teams (due to lack of planning) associated with diminished team performance. This suggests that the synthetic teams did not show enough adaptive complex behaviors that were evident in control teams. This finding implies that merely believing the pilot to be a synthetic agent made it more difficult for synthetic teams to plan and, thus, effectively anticipate their teammates' needs. In addition, this study highlights that there is a significant need for humans to gain experience in working with a synthetic agent to overcome negative perceptions.

Original languageEnglish (US)
Pages (from-to)258-267
Number of pages10
JournalIEEE Transactions on Emerging Topics in Computational Intelligence
Issue number4
StatePublished - Aug 2018


  • Autonomous agent
  • human-autonomy teams
  • nonlinear dynamical systems
  • synthetic agent
  • teamwork

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Computational Mathematics
  • Control and Optimization


Dive into the research topics of 'The Impact of Perceived Autonomous Agents on Dynamic Team Behaviors'. Together they form a unique fingerprint.

Cite this