Taken by itself, an ant does what it can: tiny size, not exceptional intelligence, and so on. And yet, you know it: when ants act as a group, they can achieve amazing results. Harvard knows this too, where researchers have developed robots capable of replicating the cooperative behavior of ants to independently solve complex problems.
Help! They will invade us!
Think before you panic. Even ants, after all, rely on algorithms. Behavioral algorithms, to be precise. In fact, each ant plays its role without thinking about the specific task, but following a series of "instincts" that guide it towards surprising engineering results, such as the creation of underground tunnels or the construction of bridges. How do ants coordinate so perfectly? They communicate with each other through antennas and leave pheromone trails to share information and orient themselves.
The research: phase one
The Harvard team began their study by involving ants in a first small-scale experiment. They placed them in a sort of circular enclosure surrounded by a soft wall of sand. Yes, a prison, you guessed right.
Initially, the ants moved at random, but soon some of them started digging into the walls in scattered spots: then, suddenly, the magic. A total strategic change.
Ants have started interacting more in specific areas. In other words, they have begun to gather in these precise "information exchange" places. This then led them to work together in one precise point and to dig a larger tunnel from which they "escaped". Collaboration, work organization, goal achieved. Okay, there biomimetics has something new to teach us.
Phase two: the Rants, robot ants, are born
From the observation of the first experiment, the researchers have begun to obtain elements for the construction of mathematical models that describe the behavior of the ants. And from the models they went on to build the Rants, robotic "ants" with the behavior of the real ones.
To communicate with each other, the Rants didn't emit chemical pheromones but left behind bright fields, or "photomonons," which grew brighter as the robots passed.
The Rants have been programmed to follow three simple rules: First, follow the gradient of the photochromic field. Second, avoid other robots where the field density was high. Third, pick up obstacles where the density was high and move them to low density areas.
Result? The rules implemented in the Rants allowed them to cooperate much like ants. When they too were put "in prison", the robots quickly realized that the best way to escape was to work together to focus on a single point of escape.
See for yourself.
Robot ants, what can they lead to?
Imagine having an army of simple but hardworking robots, capable of performing complex tasks working in perfect sync – just like a swarm of ants. Here's what could become possible thanks to this new technique.
The Harvard team authoring the research argues that this method could be used on a larger scale, involving hundreds of robots for various purposes. Moreover, such a system would make it possible to complete a job even in the event of malfunctions of some robots.
"We have shown how cooperative work can start from simple behavioral rules. A method like this can be applied to solve even very complex problems such as construction, search, rescue and defence," he says. S Ganga Prasath, co-author of the study.
The research was published in the journal eLife, I link it to you here. And be careful not to leave crumbs around.
