The EKF based Localization and Initialization Algorithms with UWB and Odometry for Indoor Applications and ROS Ecosystem

Abstract

This paper will cover some extension modules over the Turtlebot3 libraries using ultra-wideband (UWB) sensors and propose a solution to the initialization problem along with the localization problem. The Turtlebot3 already has an algorithm named move base for autonomous drive, which uses Light Detection and Ranging (LiDAR) and odometry to localize itself and avoid obstacles. However, it suffers from autonomous initialization. Therefore, ranging data from UWB sensors are used to take the initial pose of the robot to eliminate the initialization problem and advance the move base algorithm to be more robust. This data is also used in the Extended Kalman Filter (EKF) along with odometry to localize the robot. To enable wide-spread adoption, we provide an open source implementation of our algorithms and modules for the robot operating system (ROS) for real environment. Furthermore, we create an open source simulation environment for applications, which use UWB, LiDAR, and odometry data. © 2020 IEEE.