FAILURE RECOVERY PLANNING FOR ROBOTIZED ASSEMBLY BASED ON LEARNED SEMANTIC STRUCTURES

In complex task domains such as assembly and disassembly, robots need to reason about the tasks and the environment in order to make decisions. Reasoning should be based on experience acquired by the robot through interaction with users and other robots. This paper focuses on the use of learned semantic structures for failure recovery in assembly, a classical problem that, nevertheless, is far from receiving from the robotics community the due attention. A method for failure recovery planning guided by learned semantic knowledge is described, formally analyzed and empirically evaluated.