Photo: Elham Kowsari
In his doctoral dissertation, MSc (Tech) Prashant Kumar Rai investigated the use of millimeter-wave imaging radar as a primary perception sensor for autonomous systems operating in environments where vision and satellite navigation are unreliable. His work introduces learning-based radar perception methods that estimate ego-motion and spatial representations directly from high-resolution radar signal data. The study demonstrated that raw radar heatmaps can be used for consistent ego-motion estimation and place recognition without handcrafted features or point-cloud generation. Rai also developed an uncertainty-aware fusion approach that integrates radar-derived velocity estimates with inertial measurements using an adaptive extended Kalman filter. This framework can be applied to autonomous machines operating in degraded visual and GNSS-denied environments, supporting more reliable navigation in industrial and safety-critical applications.
The doctoral dissertation of M.Sc. (Tech.) Prashant Kumar Rai in the field of Automation Science and Engineering titled Towards Perception for Autonomy with 4D mmWave Radar: Learning Ego-Motion, Place Recognition, and Uncertainty-Aware Sensor Fusion will be publicly examined at the Faculty of Engineering and Natural Sciences at Tampere University on 13.02.2026.
The Opponent will be Professor Tomi Westerlund from the University of Turku. The Custos will be Professor Reza Ghabcheloo from the Faculty of Engineering and Natural Sciences at Tampere University.