This post is also available in:
עברית (Hebrew)
A recent study published in the Proceedings of the National Academy of Sciences introduces a new approach to coordinating swarms of robots—one that mimics how animals like birds, bees, and fish behave in nature. The method could improve how robotic swarms are used in real-world applications such as search-and-rescue operations, wildfire tracking, or even targeted drug delivery.
According to TechXplore, at the core of the research is a concept the scientists call “curvity”—a physical property that determines how an individual robot moves in response to external forces. Much like electric charges can be positive or negative, robots in this model are designed with either a positive or negative curvity. This determines whether they are drawn to or repelled from others in the swarm, influencing whether they form tight clusters, flow in one direction, or spread out.
Unlike traditional approaches that rely on central control or complex communication between units, this system is decentralized. Each robot operates independently based on simple geometric rules, similar to the self-organizing behavior seen in animal groups. These rules were derived using principles from natural computation and basic mechanics, making them relatively easy to implement in physical systems.
The researchers tested their theory using both simulations and small-scale robot swarms. Results showed that the curvity-based rules worked not just at the level of individual robot pairs, but scaled up effectively to thousands of units acting together.
What sets this framework apart is its simplicity. Because it’s rooted in the geometry and mechanics of how the robots are built, rather than software or external coordination, it could be adapted for use in a wide range of swarm robotics applications—whether on the ground, in the air, or even inside the human body.
The study repositions swarm control as a materials design challenge rather than a software one, opening new possibilities for robotics in both industrial and biomedical settings.