Indoor Robot Localization and Mapping Using ZigBee Radio Technology

Kyle Hevrdejs, Jacob Knoll, Advisor: Dr. Suruz Miah

Background

In the world of mobile robotics, one of the major focuses is in the area of localization and mapping. Although there are many devices currently on the market, they are often made for specific platforms and can be expensive. These two factors frequently put such systems out of the reach of researchers or hobbyists. In this project, we propose and implement a cost-effective, modular, and easily scalable framework using easy to obtain components in order to make localization and mapping easier to implement.

Project Solution

The Extended Kalman Filter Simultaneous Localization and Mapping (EKF-SLAM) is a popular algorithm used in the field of mobile robots due to its ability to perform in noisy environments. Acting as beacons for this algorithm, XBee radios are placed throughout the robots environment which provide range-only received signal strength (RSS) data. To acquire this data, a customized radio transceiver was built made up of a BeagleBone Black, stepper motor, stepper motor driver, parabolic relfector, and a XBee radio. The command signals are sent to the stepper motor driver to turn a XBee mounted at the focal point of the parabolic reflector. As the the reflector rotates, it collects RSS information from each XBee and the data is passed to the BeagleBone Black. The range and bearing of each XBee is approximated by the BeagleBone Black which is necessary for the EKF-SLAM algorithm. EKF-SLAM prediction and update steps are performed to estimate the robots position and orientation (pose), and the positions of each XBee radio. A motion control strategy consisiting of a fuzzy logic controller and  proportional controller is then performed to generate a trajectory towards the next waypoint in its path. These steps are repeated as the mobile robot navigates through the set of waypoints.