Human-inspired strategies for controlling swarm systems.

The control of swarms has emerged as a paradigmatic example of human-autonomy teaming. This review focuses on understanding human coordination behaviours, while controlling evasive autonomous agents, to inform the design of human-compatible teammates. We summarize the solutions employed by human dyads, as well as the verbal communication and division of labour strategies observed in four-person teams using virtual simulations.

Partner gaze shapes the relationship between symptoms of psychopathology and interpersonal coordination.

Interpersonal coordination is a key determinant of successful social interaction but can be disrupted when people experience symptoms related to social anxiety or autism. Effective coordination rests on individuals directing their attention towards interaction partners. Yet little is known about the impact of the attentional behaviours of the partner themselves.

Behavioral dynamics of conversation, (mis)communication and coordination in noisy environments.

During conversations people coordinate simultaneous channels of verbal and nonverbal information to hear and be heard. But the presence of background noise levels such as those found in cafes and restaurants can be a barrier to conversational success. Here, we used speech and motion-tracking to reveal the reciprocal processes people use to communicate in noisy environments.

Dynamical biomarkers in teams and other multiagent systems.

Effective team behavior in high-performance environments such as in sport and the military requires individual team members to efficiently perceive the unfolding task events, predict the actions and action intents of the other team members, and plan and execute their own actions to simultaneously accomplish individual and collective goals. To enhance team performance through effective cooperation, it is crucial to measure the situation awareness and dynamics of each team member and how they collectively impact the team's functioning. Further, to be practically useful for real-life settings, such measures must be easily obtainable from existing sensors.

Predicting and understanding human action decisions during skillful joint-action using supervised machine learning and explainable-AI.

This study investigated the utility of supervised machine learning (SML) and explainable artificial intelligence (AI) techniques for modeling and understanding human decision-making during multiagent task performance. Long short-term memory (LSTM) networks were trained to predict the target selection decisions of expert and novice players completing a multiagent herding task. The results revealed that the trained LSTM models could not only accurately predict the target selection decisions of expert and novice players but that these predictions could be made at timescales that preceded a player's conscious intent.