The purpose of this study is to simplify the motion generation during object manipulation with a multi-DOF mobile robot such as unmanned aerial vehicles (UAVs) and unmanned underwater vehicles (UUVs), thereby reducing the workload and improving usability for teleoperation.
For UAVs, as an example of object manipulation, the UAV pushed the object and moved it to the target position. Using the physical model of pushing motion, the manipulation task was completed using only the target position as a simple control command.
For UUVs, as an example of an application for object manipulation with a mobile robot, a UUV approached and collected objects floating on the water surface. A GUI that specifies the target's position and collects it was constructed using a UAV viewpoint. The subjective evaluation verified the effect of simplification of the motion generation, and the experiment using real vehicles verified the practicality of the proposed systems.