Enhancing Human-Robot Collaboration through Projected Augmented Reality and Reactive Motion Planning

The objective of the following thesis project is to explore the use of projected augmented reality and new forms of motion control applied to a collaborative robotic cell. In particular, the thesis involved the development of a native Python framework in ROS2 Humble to project useful information for the human operator onto the workspace, such as the robot's trajectory and the position of various objects. In parallel, the use of the MoveIt Hybrid Planning framework was investigated. The choice of this framework stems from its compatibility with existing configuration files, as MoveIt is already the framework used by the company. The new framework expands upon the possibilities provided by traditional motion planners by offering real-time trajectory adaptations better suited for dynamic and collaborative scenarios. The thesis work resulted in the development of custom plugins for the Hybrid Planning framework to dynamically avoid new obstacles appearing within the workspace during movement.