Specialists from the Scientific Center for Information Technologies and Artificial Intelligence at Sirius University have developed algorithms for decentralized robot control. The technology will allow coordinating groups of autonomous devices without a single command center and increase the resilience of systems to communication failures.
The approach is biomimetic: technical solutions are created by analogy with mechanisms in living nature. The next stage is studying the behavior of ant colonies.
Scientists plan to create experimental stands to find out how insects coordinate actions without centralized control, use memory, interact with each other, and change strategies depending on conditions.
We are particularly interested in how personal memory, contacts between insects, and the traces they leave together ensure the transition from foraging to regular food delivery.
The developments can be applied in unmanned aviation, search and rescue operations, agriculture, energy, and logistics. The data obtained will help create mathematical models for systems capable of independently reconfiguring collaborative work.