Understanding and Modeling Teams As Dynamical Systems

Abstract

By its very nature, much of teamwork is distributed across, and not stored within, interdependent people working toward a common goal. In this light, we advocate a systems perspective on teamwork that is based on general coordination principles that are not limited to cognitive, motor, and physiological levels of explanation within the individual. In this article, we present a framework for understanding and modeling teams as dynamical systems and review our empirical findings on teams as dynamical systems. We proceed by (a) considering the question of why study teams as dynamical systems, (b) considering the meaning of dynamical systems concepts (attractors; perturbation; synchronization; fractals) in the context of teams, (c) describe empirical studies of team coordination dynamics at the perceptual-motor, cognitive-behavioral, and cognitive-neurophysiological levels of analysis, and (d) consider the theoretical and practical implications of this approach, including new kinds of explanations of human performance and real-time analysis and performance modeling. Throughout our discussion of the topics we consider how to describe teamwork using equations and/or modeling techniques that describe the dynamics. Finally, we consider what dynamical equations and models do and do not tell us about human performance in teams and suggest future research directions in this area.

Keywords: communication analysis; interpersonal coordination; non-linear dynamics; team cognition; teams; teamwork.

FIGURE 1

(A) A task demonstrating how perceptual coupling and interpersonal interaction induces spontaneous synchronization between people; (B) relative phase of participants’ finger oscillations over a one-minute trial; (C) power spectra indicating the peak frequencies of participants’ finger movements when vision is occluded (left), un-occluded (middle), and once again occluded (right) (from Gipson et al., 2016; reprinted with permission).

FIGURE 2

In the team tying task each person handles one lace using one hand but otherwise attempts to tie a shoelace as they normally would.

FIGURE 3

The black Arnold tongues represent the periodic behavior of coupled oscillators in an iterated circle map (𝜃_n+1 = 𝜃_n + Ω - K/2π × sin[2π𝜃_n];𝜃 = phase of oscillation). The width of the Arnold tongues corresponds to predicted stability of frequency ratios as a function of the intended ratio (Ω) and coupling strength (K) between coupled oscillators (performance of the circled ratios is described in the text) (from Gorman et al., 2017; reprinted with permission).

FIGURE 4

(A) Accuracy of interpersonal coordination of mirroring (1:1) and non-mirroring (2:1–5:1) patterns aligns with Arnold tongue predictions; (B) visual occlusion (lower coupling strength) makes any ratio less stable (more error) above the critical fusion rate (1,000 ms update rate); however, humans tend to fill in missing information for any ratio when the presentation rate is below the critical fusion rate (60 ms) (from Gorman et al., 2017; reprinted with permission).

FIGURE 5

(A) A highly skilled Double Dutch team at the National Double Dutch League summer camp; (B) performance of a 7:5 (foot:rope) ratio by the team (from Gorman et al., 2017; reprinted with permission).

FIGURE 6

To maintain a stable ratio on a global (overall pattern) scale, teams can vary their patterns on a local (cycle-by-cycle) scale.

FIGURE 7

(A) The inverted pendulum; (B) UAV team coordination score; (C) short-range persistence and long-range antipersistence of the coordination score follows inverted pendulum dynamics.

FIGURE 8

(A) Code frequencies for the sample sequence of codes; (B) a simple linear transition (Markov) model of the most probable Lag-1 code transitions; (C) hypothesized temporal nesting (i.e., fractal structure) of code transitions organized around task-relevant communication.

FIGURE 9

These figures show how the knoweldge relatedness of communication diminishes as timescale (distance between utterances) increases. The communication of the experienced team in the top panel has more long-memory than the communication of the novice team in the bottom panel.

FIGURE 10

(A) Discrete recurrence plot of speaker turn-taking in a medical simulation. The black trace measures the communication determinism (larger values mean more orderly; smaller values mean more random) around the main diagonal using a moving window of size 150. (B) The black trace measures the simultaneous neurodynamic entropy across team members.

FIGURE 11

Measuring the team response to a roadblock (“relaxation time”) as a method for team assessment.

FIGURE 12

(A) Graph of communication determinism (%DET) and root mean square error (RMSE) from a prediction model; (B) RMSE relative to a 99% confidence interval (green line) indicates a significant fluctuation in communication pattern (drop in %DET) in response to the fire in the OR.