Simultaneous Localization and Mapping (SLAM) consists of a robot building a map of its surroundings using an estimate of its location while simultaneously estimating its location using the map of its surroundings. Due to the chicken-and-egg nature of this problem, there are various types of SLAM algorithms which perform well under different circumstances. We use ORB-SLAM on our quadrotors to provide odometry data, ORB-SLAM is a monocular SLAM solution which detects features and localizes based on the optical flow of these features. ORB-SLAMĀ is robust to severe motion clutter, allows wide baseline loop closing and relocalization, and includes full automatic initialization.


We are working on implementing ORB-SLAM with multiple robots sharing one map, for more efficient localization and mapping.