The MeRLIn research group of the Department of Electronics, Information and Bioengineering - Politecnico di Milano received the Best Conference Paper Award for the paper ‘Semantic Behaviour Tree Learning through Kinesthetic Demonstrations for Position-Force Controlled Robotic Applications’.
The paper was presented at the 20th IEEE International Conference on Automation Science and Engineering, held in Bari, Italy, from August 28 to September 1, 2024, and was written by Lorenzo Fratini, Niccolò Lucci, Matteo Malavenda, Andrea Zanchettin and Paolo Rocco.
The paper describes a method to improve the semantic understanding and decision-making autonomy of robots by exploiting Programming by Demonstration. The proposed method allows robots to learn and represent skills through Behaviour Trees that integrate semantic knowledge and practical execution. These trees can be further enriched to improve the robot's flexibility in performing certain actions.
By integrating Behaviour Trees with a knowledge base and a Planning Domain Definition Language planner, the robot acquires more autonomy, flexibility and the ability to adapt to unexpected events.