The main problem in creating a swarm of robots, scientists say, is overcoming and avoiding arbitrary obstacles. This usually leads to a violation of the order in which the group moves: individual devices can lag behind and get lost.
As a result, energy consumption and operating time increase, as well as its quality decreases, since there are missing sections. In the new work, YGTU specialists have created an algorithm based on the swarming behavior of animals.
Our algorithm combines two strategies: the global one, which copies the elements of the swarm intelligence, allows robots, when moving in an uncertain environment, to stay in a group and follow the target, and the local one helps them to form the correct formation, focusing on the movement of the swarm. There are no direct analogues of our development in the scientific literature today.
Oleg Maryasin, Associate Professor, Department of Cybernetics, YaGTU
The peculiarity of the new development lies in the orientation of robots using marker indicators – these are special signs that arise in the environment when performing a task.
When cleaning, they can be the moisture of the floor or traces of detergent. Scientists are confident that taking into account these signs, coupled with information from audiovisual sensors, will allow robots to act most efficiently.
As a result, the swarm is divided into groups, each of which follows its target.