We are proud to announce, that during the 29-th International Conference on Methods and Models in Automation and Robotics 2025 (MMAR 2025), Mr. Grzegorz Kaczmarczyk was awarded the Roman Kaszyński Award in the competition for the best paper delivered by a Young Author, for the work entitled "Speed Sensor Fault Detection Algorithm Based on Autoencoder Applied for Two-mass System”. The laureate is a third year doctoral student of Doctoral School of Wroclaw University of Science and Technology. His research is conducted in the Department of Electric Machines, Drives and Measurements under the supervision of Marcin Kaminski, Ph.D., D.Sc., assoc. prof.

The authors, Grzegorz Kaczmarczyk, M.Sc. (Eng) and Radosław Stanisławski M.Sc. (Eng), accomplished their research led by Marcin Kaminski, Ph.D., D.Sc., assoc. prof. It was conducted in cooperation with prof. Danton Ferreira, from Federal University of Lavras in Brazil. The authors provide the following comments on their work:

During the research we elaborated the speed sensor fault detection algorithm, applied in a two-mass electric drive system. Our main goal was to develop a reliable and efficient tool dedicated specifically to Active Disturbance Rejection Algorithm. Its main assumption says that it uses only one speed feedback coming from the load machine. Lack of direct and unambiguous correlations between available state space variables (motor current and load machine speed) creates huge difficulties in potential speed sensor fault detection using classical, analytical methods. The developed algorithm is based on the Autoencoder neural network, which allows us to detect a speed sensor fault during one numerical step. The most crucial feature of our work is the fact that we managed to obtain robustness against fake fault detection, which could potentially occur during dynamic states of the drive.