Photo: Wenyan Yang
One of the biggest challenges in artificial intelligence is Physical AI, that is, enabling robots and machines to learn and perform real-world tasks. In his doctoral dissertation, MSc Wenyan Yang developed novel robot learning methods based on imitation learning. In imitation learning, an expert, such as a human operator, provides demonstrations of how to perform a task. The robot then learns from these demonstrations and further improves its behaviour through its own exploration. Yang's methods were successfully applied to several real-world tasks, including earthmoving operations.
The doctoral dissertation titled Robot Learning from Limited Demonstrations: From Pile Loading to Scalable Generalization by MSc Wenyan Yang will be publicly examined in the Faculty of Information Technology and Communication Sciences at Tampere University on 2 October 2026. The dissertation is in the field of Signal Processing and Machine Learning.
The Opponents will be Professor Christian Smith from the KTH Royal Institute of Technology in Sweden and Associate Professor Ville Hautamäki from the University of Easter Finland. The Custos will be Professor Joni Kämäräinen Faculty of Information Technology and Communication Sciences.