Max Robotics — Global Service Robot Directory + US Certification

About

Philipp Becker is a robotics and machine learning researcher whose work sits at the intersection of robot learning, control systems, and skill-based robot programming. His research focuses on developing intelligent models that enable robots to acquire and execute complex motor skills with human-like precision and adaptability. Becker's most prominent contribution lies in hybrid inverse dynamics modeling, where his 2022 paper on end-to-end learning of hybrid inverse dynamics models (11 citations) demonstrated how combining principled physics-based components with data-driven approaches can significantly enhance tracking performance and compliant impedance control in robotic systems. This work addresses long-standing challenges such as modeling stick-slip friction — phenomena that purely analytical models struggle to capture. Beyond dynamics modeling, Becker has made meaningful contributions to skill libraries and versatile robot behavior, exploring how robots can specialize reusable skills across varied tasks, as seen in his work on local mixture of experts. His more recent research on the Multimodal Trajectory Transformer (MuTT) pushes toward reducing the manual effort required when configuring robot skill parameters, leveraging multimodal learning to streamline real-world deployment. Across these contributions, Becker exemplifies a rigorous, systems-minded approach to making robots more capable, adaptable, and practically deployable.