Max Robotics — Global Service Robot Directory + US Certification

About

Ilija Radosavovic is a pioneering researcher at the forefront of robot learning, embodied AI, and autonomous systems. His work centers on enabling robots to operate effectively in real-world, unstructured environments through advances in reinforcement learning, self-supervised pre-training, and large-scale data collection. Radosavovic's most celebrated contribution is his work on humanoid locomotion using reinforcement learning, demonstrating that humanoid robots can navigate diverse real-world terrains with remarkable autonomy — a paper that has already accumulated over 150 citations since its 2024 publication. He has been a key contributor to the Open X-Embodiment initiative, a landmark collaborative effort to consolidate large-scale robotic learning datasets and foundation models for robotics, drawing direct inspiration from transformative trends in NLP and computer vision. His involvement in the DROID dataset further underscores his commitment to building the data infrastructure that will power next-generation robot policies. Earlier work on masked visual pre-training established that self-supervised representations learned from natural images transfer effectively to motor control tasks, laying important groundwork for scalable robot learning. Collectively, Radosavovic's research is shaping a future where versatile, generalizable robots learn from internet-scale data and operate seamlessly alongside humans.