Slam-Bot: Seeing and Mapping

Localization and mapping are fundamental for any robot navigating in unknown environments. There are a variety of applications for the technology including autonomous vehicles, autonomous aerial vehicles, robot vacuum cleaners, toys like the Anki Drive, and industrial robots. This project is a low-budget simultaneous localization and mapping (SLAM) robot which is able to solve the problem of localization to know its current position, as well as mapping to build and update the map in real time. We apply Microsoft Kinect for Xbox 360 as the sensor of our SLAM robot for its acceptable accuracy and low cost, and we use Arduino to control the movement of the robot. An Android phone with an inertial measurement unit (IMU) app is fixed on the robot to improve the odometry, since that from the wheel encoders is prone to error from voltage changes, wheel diameter, track width miscalculations, etc.  We also apply several software libraries and tools from Robot Operating System (ROS) such as the gmapping package, which produces a 2D map from laser scan data, packages from the ROS navigation stack, and packages for the Kinect sensor to work in this project. The project offers a low-cost option. One of the applications of our SLAM robot is to map complicated indoor environments for further extensions such as indoor navigation and floor planning.