The multicriteria task of preflight and operational planning of group actions of unmanned aerial vehicles (UAVs), taking into account the required service schedule, is considered. A minimax criterion for the operational planning of group actions when the dynamic situation changes is proposed. The shape of the expert system for controlling the duration of observation during the search and detection of ground objects is formed. The obtained results of assessing the quality of the solution to the subproblem of neural network recognition of mobile objects based on deep learning confirm the effectiveness of the proposed approach in monitoring the controlled area.