The main focus of the cloudSLAM project is to develop algorithms for cooperatively solving the SLAM problem such that a robot, or a team of robots and humans can operate reliably in novel, highly dynamic environments. The cooperation is manifested via operation of a multi-agent fleet, e.g. unmanned ground vehicles (UGV), unmanned aerial vehicles (UAV), humans, etc. Each agent is equipped with a perceptive sensor (stereo camera or lidar) and an inertial measurement unit (IMU). The team is further boosted by cloud technologies, such that the map and associated uncertainties are shared in cloud, thus improving the overall efficiency of single- vs. multi-agent SLAM.



The project consists of three main parts each assigned to a different workpackage:

  • WP1. 6DoF Detection and tracking of moving objects
  • WP2. Cooperative cloud based 6DoF dynamic SLAM
  • WP3. System integration and validation

Detection and tracking of moving objects developed within WP1 will ensure high precision of 6DoF SLAM algorithm developed within WP2 regardless of the number of moving objects in the environment. All developed algorithms will be connected and tested under real world environemnt within WP3.