Communication Dynamics and Team Performance in Multiple-Operator-Multiple-Agent (MOMA) Team

The majority of existing research on human-agent teaming has focused on dyadic teams consisting of one human and one intelligent agent (IA). To address the limited research on nondyadic teams, this study investigated communication dynamics and team performance in a Multiple-Operator-Multiple-Agent (MOMA) team where two human players collaborated with two IAs to jointly deliver a sequence of colored blocks in a particular order as fast as possible. Human players were responsible for searching rooms to identify the locations of specific color blocks and issuing commands to the IAs, which could then move and pick up the blocks. The study manipulated IA reliability paring across three conditions: perfect-perfect (both IAs with 100 % reliability), perfect-imperfect (one IA with 100 % reliability and the other with 50 %), and imperfect-imperfect (both IAs with 50 % reliability). Each team completed 10 trials under each condition, resulting in a total of 30 trials. The results revealed a significant effect of the IA reliability pairing on average task completion time. Additionally, for average communication frequency, there were significant effects of the recipient (i.e., receiver of the message), reliability pairing, and their interaction. The number of commands humans issued to an IA varied depending on the reliability level of the IA it was paired with. Analyses of message types sent by humans further indicated that players adjusted their strategies based on IA reliability levels. These findings provide valuable insights into the dynamics of MOMA teams and human adaptability in collaborative settings.