The paper presents the results of the development of models, algorithms and software for planning the routes of a truck and a group of autonomous quadcopters in an urban environment for delivering goods to points of sale. A virtual model of the urban environment and autonomous vehicles as part of a heterogeneous delivery system has been developed. A quadcopter routing algorithm is proposed based on the joint application of clustering algorithms to identify groups of target points and evolutionary programming methods. The description of the approach to the organization of the automatic landing of the quadcopter in the target position is also given. The results of experimental studies on evaluating the effectiveness of solving the problem of delivering goods with specified parameters of the routing algorithm for a system consisting of one cargo vehicle and three autonomous quadrocopters are described and analyzed