(EU Chist-Era; 2019-2022)
This project will provide scientific advancements for benchmarking, object recognition, manipulation, learning and human-robot interaction. Our problem is to sort a complex, unstructured heap of unknown objects –resembling nuclear waste consisting of a set of broken deformed bodies– as an instance of an extremely complex manipulation task. The ability of robots to fully autonomously handle dense clutters or a heap of unknown objects has been very limited due to challenges in scene understanding, grasping, and decision making. Here, we will rely on semi-autonomous approaches where a human operator can interact with the system (e.g. using tele-operation but not only) and giving high-level commands to complement the autonomous skill execution. The amount of autonomy of our system will be adapted to the complexity of the situation. We develop a manipulation skill learning system that learns from demonstrations and corrections of the human operator and can therefore learn complex manipulations in a data-efficient manner. To improve object recognition and segmentation in cluttered heaps, we will develop new perception algorithms and investigate interactive perception in order to improve the robot’s understanding of the scene in terms of object instances, categories and properties.
Serena Ivaldi, Jean-Baptiste Mouret, Sylvain Lefevre, Mihai Andries (postdoc), Yoann Fleytoux (PhD student), Lorenzo Vianello (master student 2019)
- University of Lincoln (coordinator)
- University of Wien