Market Assessment for an innovative method to self-calibrate manipulator kinematic and sensor extrinsic**parameters

点击翻译按钮获取中文摘要

Collaborative robots, or cobots, are machines designed to work alongside humans in shared spaces. Many**cobots are mobile, and incorporate lightweight, person-safe robotic arms for manipulation tasks. These**(mobile) manipulators need to be correctly calibrated to ensure accurate and reliable operation. Calibration**typically involved determining sensor extrinsic parameters (i.e., external sensor pose relative to the manipulator**base) and manipulator kinematic parameters (e.g., joint biases and link length offsets). Calibration parameters**will often change over a robot's lifetime due to, for example, general wear and tear.**We have developed an innovative method for manipulator self-calibration using contact sensing. The method,**based on point cloud registration, can be applied to determine sensor extrinsic and manipulator kinematic**parameters. Requisite point cloud maps are generated from external visual or depth camera data, and by**monitoring end-effector contact with surfaces in the environment using a force-torque or tactile sensor and an**impedance controller. The process uses on-board hardware exclusively and does not rely on any external**measurement devices, fiducial markers, or calibration rigs; further, it is fully automatic in the general case.**This self-calibration method will very substantially reduce the costs of deploying person-safe, collaborative**manipulator systems, by decreasing or eliminating the need for human-in-the-loop maintenance and**adjustment. A wide range of robot application domains, from manufacturing to services, will benefit from**deploying the technology. This proposal seeks funding for a market assessment, to identify opportunities for**commercialization (within Canada), improving Canadian competitiveness in an emerging, critical market**sector with tremendous growth potential.