Researchers at Harvard University recently took inspiration from ants and designed a group of relatively simple robots that can work together to complete complex tasks using only a few basic parameters. The research was published in eLife.
Lola England de Valpine is a professor of applied mathematics and senior author of the paper. "This project has been dedicated to understanding the collective dynamics of social insects such as termites and bees, particularly how they use their environment to create complex functional architecture," he said.
The research team first studied how black carpenter ants work together to dig out and escape from soft corrals.
S Ganga Prasath is a postdoctoral researcher at the John A. Paulson School of Engineering and Applied Sciences at Harvard University and one of the lead authors of the paper. "At first, the ants in the corral move around randomly, and they communicate with each other through their antennae before starting to cooperate," he said.
Ants interact with their surroundings and other ants primarily through their antennae. Researchers found that ants spontaneously swarm around areas where they interact frequently. Once a few ants start digging tunnels in a corral, they are quickly joined by others. Over time, the tunnels become wider and more complex, and eventually the ants are able to escape.
Motivated by this understanding and model, the researchers built robotic ants nicknamed RAnts to see if they could work together to escape similar corrals. Rather than using chemical pheromones, RAnt uses "photohormones," which are light fields left behind by flowing RAnt that mimic a pheromone field or antennae.
RAnts is programmed with only simple local rules: follow the gradient of the photoreceptor field, avoid other robots with a high density of photoreceptors , and pick up obstacles where the density of photoreceptors is high and throw them to places with a low density of photoreceptors.
These three rules allow RAnts to break free quickly and, just as importantly, allow researchers to explore areas of behavior that are difficult to detect with real ants. The
RAnts robot successfully imitated ants. de Valpine said: "This work highlights the potential of designing simple and effective multi-robot systems that can complete complex tasks through local interactions. It is also a proof of concept for using simple animal behaviors as a basis for designing robots to complete complex tasks."
The successful results of these robots show that it is feasible to use simple animal behaviors as models to design robots, which helps provide more options for designing new behaviors for robots. Furthermore, the success of these robots demonstrates the effectiveness of using local interactions to complete complex tasks.
The researchers also said that the design methods of these robots could be used in other fields, such as medicine or agriculture, to help complete complex tasks. They plan to continue researching these robots and apply them to more areas in the future.
hopes to develop robots that can work together with humans in the future to help humans complete more complex tasks. For example, in hospitals, robots can assist doctors in surgery, or in factories, robots can assist workers in production processes.
