Max Robotics — Global Service Robot Directory + US Certification

About

Christian Plagemann is a machine learning researcher whose work sits at the intersection of probabilistic modeling, robotics, and spatial perception. His most influential contributions center on Gaussian process (GP) methods, particularly advancing their applicability to real-world settings through innovations in heteroscedastic and nonstationary regression. His 2007 paper on heteroscedastic Gaussian process regression, which models input-dependent noise using a second GP, has accumulated over 313 citations and remains a foundational reference in the field. Building on this, Plagemann extended GP frameworks to handle nonstationary kernels and sparse approximations, enabling efficient terrain modeling for legged robots and gas distribution mapping in dynamic environments. Beyond statistical methodology, Plagemann made notable contributions to mobile robotics, developing probabilistic sensor models for RFID-based localization and Gaussian beam processes for range finders, both of which strengthened the reliability of robot perception systems. His work on unsupervised object class discovery from 3D range data and body schema learning for manipulators reflects a broader ambition toward autonomous, self-supervising robots. Collectively, his research has garnered over 1,000 citations, demonstrating sustained influence across probabilistic machine learning and autonomous systems communities and making his methods essential reading for researchers in robot learning and spatial modeling.