Gesture-based Contactless Control of Mobile Manipulators using Capacitive Sensing

Contactless control of serial and mobile manipulators is of interest in highly sensitive environments such as clean rooms and operational rooms to circumvent contamination of surrounding materials and in collaborative robotics to ensure safe and intuitive operation on shared workspaces. We present a contactless control scheme based on capacitive sensing which enables an intuitive control of robot manipulators. Contrary to optical and vision-based systems the capacitive sensor is robust against mechanical impact, dirt and does not suffer from occlusions or bad light conditions. The sensor can be realized on a flexible substrate, which offers a variety of placement options for the sensors, e.g. directly on a robot arm or integrated in the surface of a table or workplace. A comparatively simple model based approach is used to detect gestures thus avoiding the need for large training sets and allowing for easy adaptability to various geometric constraints. The capabilities of the proposed system are demonstrated by controlling the end-effector velocity of a mobile manipulator in 3D task space combined with a visualization of the system as feedback for the operator.