Harvard Researchers Develop Robotic Ants for Building and Dismantling Structures

Researchers at Harvard have developed a fleet of robotic ants designed to replicate the self-organizing behavior of social insects. These robots, dubbed “RAnts”, build and dismantle structures without blueprints or central leadership. The work was conducted by researchers from the John A.

Paulson School of Engineering and Applied Sciences (SEAS). The robotic ants communicate using light fields called photormones, which serve as a substitute for chemical pheromones used by real ants. This system allows the robots to coordinate through environmental cues in a feedback loop.

The approach is based on stigmergy, where agents respond to changes made by others in their surroundings.

The robots follow simple rules, such as tracking light gradients, transporting blocks, and depositing them at specific thresholds. By adjusting parameters like light-following intensity and block-handling settings, the swarm can switch between construction and demolition tasks.

This enables the completion of complex tasks through decentralized, self-organized interactions. Professor L. Mahadevan stated that the study shows how simple local rules can lead to emergent complex task completion that is robust and adaptive. He also introduced the concept of exbodied intelligence, where collective cognition emerges from interactions between agents and their evolving environment.

The findings were published in the journal PRX Life.

The technology demonstrates that sophisticated tasks can be achieved with minimal individual intelligence, relying instead on group dynamics. Possible uses include autonomous construction in hazardous areas, planetary exploration, and models for studying animal behavior. The system operates without central control, making it adaptable to changing conditions.