SLAM (simultaneous localization and mapping) is the technique used by most autonomous robot, such as Roomba, to build the map in an unknown environment while navigating simultaneously. There are many existing platforms and packages to conduct SLAM, such as MATLAB, R, etc. Normally, camera, Lidar and IMU are used as the sensors.   Currently, we are implementing the SLAM algorithm in C++ on ROS and Linux with RealSense D435i.