Robotic ants collectively pull-off prison escape with no plan
Researchers, led by senior author Professor L. Mahadevan, drew inspiration from ants to design simple robots that work collectively to perform complex tasks with simple parameters. The study, published in eLife, sought to understand how individual social insects such as ants, termites and bees, work collectively to perform really complex tasks including construction, foraging, and defence.
The researchers examined how black carpenter ants work together to escape from a soft corral. Ants primarily rely on their antennae to interact with the environment and other ants, a process called antennation. The researchers observed the ants spontaneously congregating around areas where they interacted more often. Once a few ants began to tunnel into the corral, other ants joined in. From observations the team identified two important parameters, the strength of collective cooperation and the rate of evacuation. Mathematical models showed that the ants were only successful when they coorperated with each other.
WIth this understanding, the team built the robotic ants, nicknamed RAnts, to see if they could work together to escape a similar corral. The RAnts used “photormones,” fields of light that are left behind by the roving RAnts that mimic pheromone fields or antennation. The RAnts were programmed with simple rules to follow the photoromone field, avoid other robots where photoromone density was high, and pick up obstacles where photoromone density was high and drop them where photoromone was low. The researchers showed that a cooperative completion of tasks arises from simple rules and that this can be applied to solve other complex problems.
This approach can be scaled up to teams of dozens or hundreds using a range of various communication fields. Further, it's more resilient than other approaches as even if a few individual robots fail, the rest of the team can complete the task.