Max Robotics — Global Service Robot Directory + US Certification

About

Rohan Bandaru is a robotics researcher whose work sits at the compelling intersection of deep learning and physics-based optimization — a frontier that addresses one of the most persistent challenges in modern robotics: building systems that generalize reliably across dynamic, real-world environments. His most notable contribution is **PyPose**, an open-source library designed to bridge the gap between data-driven neural approaches and classical physics-based optimization for robot learning. Recognizing that deep learning excels at complex perception tasks yet struggles with environmental generalization, while physics-based methods offer robustness but lack representational power, Bandaru and his collaborators engineered PyPose as a unified framework enabling researchers to harness the strengths of both paradigms simultaneously. The work has garnered significant academic attention, accumulating 38 citations across its 2022 and 2023 publications — a strong indicator of its practical relevance and adoption within the robotics community. PyPose reflects a broader vision of making principled, physics-aware robot learning more accessible and scalable, positioning Bandaru as an emerging voice in the effort to develop more adaptable, intelligent robotic systems capable of operating reliably beyond controlled laboratory settings.