Muhammad Asif Rana
Papers
9
Total Citations
107
H-Index
7
About
Muhammad Asif Rana is a robotics researcher whose work sits at the intersection of geometric mechanics, learning from demonstration, and policy design. His most significant contributions center on developing mathematically principled frameworks for robot motion generation. Rana is the lead author of "Geometric Fabrics" (2022, 29 citations), which generalizes classical mechanics to create a unified theory for designing robot behaviors with inherent stability and physical interpretability. This work extends his earlier foundational contributions to Riemannian Motion Policies (RMPs) and the "Euclideanizing Flows" approach (2020, 16 citations), which learns stable dynamical systems from limited human demonstrations by diffeomorphically transforming complex motions into simpler spaces. His research also includes "Generalized Cylinders" (2017, 14 citations) for skill representation and "RMP2" (2021, 11 citations), a structured policy class for robot learning. Across these contributions, Rana has developed a coherent geometric framework that enables robots to learn, generalize, and refine complex skills while maintaining mathematical guarantees on stability and safety. His work bridges classical mechanics, differential geometry, and modern machine learning, offering rigorous tools for designing robust, interpretable robotic motion policies.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Geometric Fabrics for the Acceleration-based Design of Robotic Motion14 citations · 2020
- 5RMP2: A Structured Composable Policy Class for Robot Learning11 citations · 2021
- 6
- 7
- 8
- 9RMP2: A Structured Composable Policy Class for Robot Learning3 citations · 2021