ICP-EKF localization with adaptive covariance for a boiler inspection robot

The boiler inspection robot was developed for inspecting the thickness of a pipe wall in a boiler with an electromagnetic acoustic transducer (EMAT) probe and cameras. The robot needs to be localized during the inspection process to correlate the measured data with the inspected location. The localization technique uses an iterative closest point matching (ICP) algorithm together with an extended Kalman filter (EKF). Artificial landmarks were placed in the environment to help the localization process. The covariance of the process noise and the measurement noise were automatically adjusted based upon the command input and the number of landmarks detected by the robot. The experimental results showed that the proposed adaptive covariance can help improve the localization performance of the robot on the pipe wall.