Determination of the target configuration of a mobile manipulator in tasks of grasping objects

The task of collecting and moving objects using a mobile manipulator (a mobile robotic platform equipped with a robotic arm) is being considered. To solve this task, it is necessary to determine the target configuration of the mobile manipulator for grasping the target object, taking into account the position of this object and obstacles. The target configuration includes the position of the mobile platform and the configuration of the manipulator’s links. An algorithm for determining the target configuration of a mobile manipulator in a confined environment, where a high density of obstacles requires considering the geometric shape of the wheelbase and the workspace of the manipulator during grasp planning, has been proposed and implemented. The algorithm performs a series of obstacle map convolutions with kernels formed considering the shape of the base and the manipulator’s workspace, resulting in a set of permissible platform positions from which the manipulator can grasp the object. During experiments in the Isaac Sim simulation environment, the proposed approach demonstrated its effectiveness for confined spaces.