dc.description.abstract | The master’s thesis project explores Leo Rover, equipped with a Lidar sensor, stereo camera, and robotic arm, available at University of South-Easten Norway. Leo Rover is a semi-autonomous mobile robot used for research and development purposes using the robot operating system (ROS).
The project aims to develop a computer vision-guided autonomous grasping system, where the rover detects nearby AR-tags (fiducial markers) based objects, determines the most suitable grasping approaches using its robotic arm, plan and execute Motion Planning (MoveIt) to pick-up and place small objects. The rover also performs autonomous navigation, including simultaneous localization and mapping (SLAM). Various methods are used including utilization of OpenCV tool for real-time object detection, configuring and integration of rover’s ROS network, testing multiple ROS packages, evaluating rover and its arm’s performance in both simulated and physical environment.
The thesis report provides valuable insights into the capabilities and limitations of the rover and its robotic arm. Consequently, a comprehensive risk assessment is carried out to identify potential safety concerns and ethical factors. This research serves as a foundational setup for future work for autonomous navigation and advanced grasping technologies in mobile robotics. | |