To identify a swarm of robots, it is necessary to teach them to analyze a set of practical tasks – but at the same time, neither to analyze situations nor the result depends on the state of individual individuals. Simply put, you can appoint commanders and foremen among the robots to manage the strategy, since if they are implemented or made a mistake, the work will fail. In search of common solutions from Harvard, they found a common option – the ant interaction system.
A worker ant knows how to work, but by itself does not determine where and what to apply its efforts to. No, it is necessary, because his primary task is to work about the anthill, and not about the need for needs. Therefore, the ants are constantly in contact with each other with the help of antennae and follow the trails of pheromones to find the place of the accumulation of other ants. Where everyone has gathered is where the most important thing is at the time of getting a job.
In the experiment, the scientists placed a group of ants in a sand-walled cell trap. Some ants spontaneously observed it in different places, but they were few. However, when the ants gathered in one place, they easily dug out a way out. Based on the experiment, the initial “robouravies” were created that could crawl, move grain simulators and leave a light trail on the surface.
The robot ants received only three instructions. First, do not collide with each other. Second, drag the grains of sand from the economy with a bright light trail to a dark one. Thirdly, follow yourself out of the dark areas last. After some time, robouravies, without noticing it, successfully dismantled the barrier and happened to be free. This aorhythm may be useful for specific programs.
Source: Tech Cult
