A distributed multi robot SLAM system for environment learning

This paper presents a multi mobile robot simultaneous localization and mapping (SLAM) system for feature based environment learning by using team of exploring robots. The environmental information is measured through the dynamic sensor network in the shape of moving robots with unknown initial poses. Each robot is equipped with 2D laser scanner and a webcam and it serves as a moving sensor node to perceive horizontal and vertical line features respectively. All the moving nodes are responsible to build the informational structured space. The proposed system is using a unified Extended Kalman Filter (EKF) based SLAM framework for each robot which eventually builds a line feature based partial 3D model of the environment. Each moving robotic sensor node then shares its feature based map model to other moving nodes which are in communication range. A global map model is then transformed after getting mutual pose estimation of the robots by matching mutual common map features in addition by taking visual confirmation of other robot. The proposed system has been tested in an indoor environment and results are shown in the paper.