Nicholas Rotella

University of Southern California, Robotics Research (United States)

Papers

Total Citations

484

H-Index

About

Nicholas Rotella is a leading researcher in humanoid robotics, specializing in whole-body control, state estimation, and momentum-based manipulation for dynamic locomotion. His most influential work, "Momentum control with hierarchical inverse dynamics on a torque-controlled humanoid" (254 citations), established a foundational framework for managing a robot's overall momentum through prioritized task execution—a critical advance for stable, agile movement in complex environments. Rotella further transformed the field with his state estimation framework (106 citations), which enables humanoid robots to accurately determine their position and orientation using only onboard proprioceptive sensors and leg kinematics, eliminating the need for external motion capture. His research on multi-contact momentum control (88 citations) extended simplified dynamic models like the Linear Inverted Pendulum Model to handle non-coplanar contacts, allowing robots to navigate stairs and uneven terrain. More recently, Rotella has pioneered the use of torque measurement in centroidal state estimation (2023), demonstrating how drive-system torque data can enhance estimation accuracy on modern legged platforms. With over 480 total citations, his work directly enables torque-controlled humanoids to perform robust, real-world locomotion and manipulation tasks.