Summary: Researchers report that physical coordination through touch becomes more effective as group size increases. These results may inform group-based physical rehabilitation strategies and the design of multi-agent robotic systems that must coordinate movement.
Source: Tokyo Institute of Technology
Physical Coordination Improves in Larger Groups, Study Finds
Teams of researchers from Tokyo Institute of Technology, Imperial College London and the University of Tokyo have shown that physical coordination among people improves as group size grows. Using controlled experiments with robotic interfaces, the team compared pairs, trios and quartets who were physically coupled by their right hands and asked to follow a common moving target. Rather than degrading with added participants, performance steadily improved with each additional group member.
The researchers designed the task to isolate haptic communication—information conveyed through touch—so they could observe how people use interaction forces to infer the group’s collective intent. Their experiments revealed that individuals rapidly estimate the group’s goal by sensing the forces applied by others. This ability allows even highly skilled individuals to coordinate effectively with less skilled partners and produces benefits that scale with the number of participants.
Lead author Dr. Atsushi Takagi, based at the Institute of Innovative Research at Tokyo Institute of Technology, described the surprising speed and robustness of the effect. He noted that when people are physically connected, coordination by touch can emerge within seconds and reach consensus in roughly the same amount of time regardless of the group’s size. He compared it to trying to move a crowded banquet table by voice alone: verbal communication often slows coordination as more people are involved, but haptic interaction bypasses that bottleneck.
Professor Etienne Burdet of Imperial College, the study’s senior author, emphasized the counterintuitive nature of the findings. Although each additional person could introduce noisy forces, overall performance improved because each participant effectively connects to a “virtual collective hand” — a behavioral average of all partners. Averaging reduces noise, so the group signal becomes cleaner and individual members can plan movements more accurately.
Mechanism: Inferring the Collective Goal Through Touch
The core mechanism identified by the team is goal inference: each person uses the forces they feel from others to estimate the group’s intended trajectory. With that estimate, individuals adjust their own movement planning to align with the collective goal. The investigators also ran computational simulations to probe what information is transmitted through touch and how it is used to shape behavior. Simulations supported the experimental conclusion that inferring the collective target is sufficient to explain fast, scalable coordination.
Previous work by members of this collaboration showed that haptic communication in pairs enhances learning and task performance in partner-like tasks such as coordinated movement and partner dancing. The present study extends those insights to larger groups, demonstrating that the same haptic processes apply to trios and quartets and are likely critical for many social and cooperative physical activities.
Implications for Rehabilitation and Robotics
These findings have practical implications. In rehabilitation contexts, training patients in small groups could exploit haptic interaction to accelerate recovery of coordinated movement skills, potentially improving outcomes compared with isolated practice. In robotics, a clearer understanding of how humans infer collective goals through touch could inform algorithms that enable groups of robots to carry out shared physical tasks more effectively and robustly.
Research partners: Tokyo Institute of Technology, Imperial College London, University of Tokyo
Publisher of the press release: NeuroscienceNews.com (organized coverage)
Image credit: Pexels
Original research article: “Individuals physically interacting in a group rapidly coordinate their movement by estimating the collective goal” by Atsushi Takagi, Masaya Hirashima, Daichi Nozaki, and Etienne Burdet, published in eLife. DOI: 10.7554/eLife.41328. Published February 12, 2019.
Abstract
How can a human collective coordinate—for example, to move a banquet table—when each person is influenced by the inertia and actions of others who may perform the task less well? The authors tested whether larger groups would suffer a zero-sum outcome in which inferior performers hinder superior ones by examining dyads, triads and tetrads physically coupled by the right hand while following a moving target. Contrary to the hypothesis that adding individuals increases disruptive noise, superior individuals maintained accurate tracking even when coupled to inferior partners, and group interaction benefits increased with group size. A computational model indicates these benefits arise because each individual uses the interaction forces they feel to infer the collective target and improve movement planning. As a result, coordination becomes rapid, beneficial and scalable: individuals estimating the collective goal make physical interaction efficient across group sizes.