Nature-Inspired 'Link-Bots' for Self-Organizing Swarm Robotics

Link-bots exhibit collective behavior and can perform a variety of tasks, including transport of objects.  Credit- Harvard John A. Paulson School of Engineering and Applied Sciences

Inspired by the collective behavior of ant colonies and schools of fish, researchers at the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) have unveiled a groundbreaking robotic system that mimics nature’s cooperative swarming abilities—without the need for motors, sensors, or centralized control.

In a new study published in Science Advances, an international team led by Professor L. Mahadevan of SEAS and in the Department of Organismic and Evolutionary Biology, and Professor Ho-Young Kim of Seoul National University introduced “link-bots”—centimeter-scale, 3D-printed, V-shaped chains that move and work together using only vibration and simple mechanical linkages.

Each bot is composed of passive, notched links that, when placed on a vibrating surface, generate coordinated, life-like movement. The bots can collectively navigate obstacles, transport objects, and respond to environmental constraints—demonstrating behaviors usually seen only in complex, sensor-laden robots or living organisms.

“What’s remarkable is that these simple mechanical agents produce sophisticated behavior just through physical interactions,” said Mahadevan. “It’s a powerful demonstration of how complexity can emerge from simplicity.”

The research marks a step forward in low-power, scalable swarm robotics, with potential applications in material transport, autonomous sorting, and environmental monitoring.

“This shows a new design paradigm—where intelligence emerges from geometry and interaction rather than complex computation,” added SEAS postdoctoral fellow Kimberly Bowal, who developed a model to simulate and predict the bots’ behavior.

The work reimagines robotic function through the lens of self-organization, drawing parallels to the principles of evolution and natural selection.

Co-authors of the study include Kyungmin Son and Kwanwoo Kim from Seoul National University.