Localization Of A Miniature Spherical Rolling Robot Using Imu, Odometry And Uwb

Robots that rolls with a spherical body or spherical robots, exhibits a high degree of mobility and amazing recovery capability from collisions while traversing in the environment. However, the localization of spherical robots in a GPS-denied environment for Intelligent Surveillance and Reconnaissance (ISR) task is a challenging problem due to the complexity of its system dynamics and the limited available sensors technology to sense out of the spherical shell. In our prior work, a kinematic localization technique based on odometry and inertial measurement unit (IMU) sensing was proposed and implemented onto our miniature spherical robot Virgo, for pose estimation. However, it suffers from errors due to slippages during locomotion or as a result of the collision. In this paper, we present a solution to this problem by the inclusion of an additional ultra-wideband (UWB) sensor and fuse it with our kinematic pose estimator using Extended Kalman Filter for indoor localization. Experiments are conducted on a multi-waypoint trajectory to verify its validity and had shown to improve localization performance.