Papers
11
Total Citations
119
H-Index
6
About
Anqi Li is a robotics and machine learning researcher whose work sits at the intersection of geometric control theory, motion planning, and multi-robot systems. She has made significant contributions to the development of **Riemannian Motion Policies (RMPs)** and **geometric fabrics** — mathematically principled frameworks that extend classical mechanics to enable robots to navigate complex environments while satisfying multiple objectives simultaneously. Her 2022 paper on Geometric Fabrics (29 citations) advances the theoretical foundations of behavior-driven robot control, while her work on Euclideanizing Flows (16 citations) demonstrates how diffeomorphic transformations can enable robots to learn stable, smooth motions from limited human demonstrations. Li has consistently tackled the challenge of multi-objective task composition, developing provably stable frameworks for combining concurrent control policies across single and multi-robot settings. Her RMP2 framework further bridges structured geometric control with modern machine learning, enabling robots to acquire motion policies from demonstration data. With contributions spanning theoretical foundations, practical robot design, and learning-based methods, Li's research provides a coherent and growing body of work that is shaping how robots reason about motion in geometrically rich environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Geometric Fabrics for the Acceleration-based Design of Robotic Motion14 citations · 2020
- 5Stable, Concurrent Controller Composition for Multi-Objective Robotic Tasks13 citations · 2019
- 6RMP2: A Structured Composable Policy Class for Robot Learning11 citations · 2021
- 7A Sequential Composition Framework for Coordinating Multirobot Behaviors6 citations · 2020
- 8
- 9
- 10RMP2: A Structured Composable Policy Class for Robot Learning3 citations · 2021