Max Robotics — Global Service Robot Directory + US Certification

About

Michiel Blokzijl is at the forefront of embodied AI, pioneering the development of generalist agents that can learn and adapt across diverse robotic platforms. His most impactful work introduces **RoboCat**, a self-improving agent that leverages heterogeneous robotic experience to rapidly master novel skills and embodiments, marking a paradigm shift away from task-specific models toward truly versatile robotic intelligence. With 9 citations, this foundational paper demonstrates how diverse training data can unlock unprecedented adaptability in physical agents. Building on this, Blokzijl’s recent **Gemini Robotics** report (2025) tackles the grand challenge of translating large multimodal models from digital domains into the physical world, purposefully designing AI architectures for real-world robotic interaction. His research sits at the critical intersection of foundation models and robotics, addressing how generalist digital capabilities can be grounded in physical action. Blokzijl’s work is shaping the next generation of robots that learn continuously, adapt to new tasks with minimal data, and operate across multiple embodiments—a vision that promises to democratize robotic manipulation and accelerate the deployment of intelligent machines in homes, factories, and beyond.