Coordinated and autonomous movement of unmanned ground and aerial vehicles at indoor and outdoor environments

In this study, we present autonomous exploration in unmapped environments and aim to achieve complete topological coverage without relying on prior maps. Many existing approaches assume pre-built environmental models or target sensor coverage rather than topology, so they fall short of this goal. We propose two Generalized Voronoi Diagram (GVD)–based strategies that achieve coverage by systematically visiting Voronoi vertices and traversing edges. The first uses a range-sensing ground robot and maintains equidistance from nearby obstacles to navigate precisely. The second forms a cooperative UAV–UGV team: a camera-equipped aerial vehicle guides the ground platform and enables predictive decisions. We implement both strategies in a unified co-simulation pipeline that integrates MATLAB/Simulink with Gazebo and includes realistic vehicle dynamics and sensor models. We evaluated the methods in four environments with topological complexity increases from 8 to 39 Voronoi vertices. The results revealed a clear trade-off between spatial accuracy and operational efficiency. The range-sensor method spatial achieved 3-6 times higher precision through geometric vertex computation. By contrast, the aerial-camera method completed missions two to three times faster and improved path efficiency by 17–21% through predictive navigation.