Foo Bar Identity
Foo Bar design system including logo mark, website design, and branding applications.
Posts by Collection
portfolio
Ginger Gulp Identity
Ginger Gulp design system including logo mark, website design, and branding applications.
Impact-Aware Task-Space Quadratic-Programming Control
We enable swift robot motion without fearing impacts with the environment. For instance in a box manipulation task, the robot does not need to stop or use cl...
speakers
Alphonsus Adu-Bredu
Exploring Kinodynamic Fabrics for Fast, Reactive Whole-Body Control of Underactuated Humanoid Robots
Sven Behnke
Direct Centroidal Control for Balanced Humanoid Locomotion
Bike Zhang
Learning Humanoid Locomotion via Transformer
Carlos Mastalli
Inverse-Dynamics Model Predictive Control
Maurice Fallon
Sensor fusion based legged robot perception, state estimation and exoskeleton navigation
Shi Fan
Reference-based vs Reference-free Learning Bipedal Motor Skills via Assistive Force Curricula
Grzegorz Ficht
Direct Centroidal Control for Balanced Humanoid Locomotion
Serena Ivaldi
Anticipatory control: using prediction of intended movement for control
Johannes Englsberger
Torque-based Dynamic Walking on Compliant and Uneven Terrain
Luis Sentis
From Model-Based Whole-Body Control to Humanoid Legged Manipulation using Machine Learning
Yan Gu
State Estimation and Control of Underactuated Humanoid Walking on a Nonstationary Surface
Tomomichi Sugihara
Model predictive control based on 3D ZMP manipulation
topics
Angular Momentum Tracking
Angular momentum (AM) tracking can improve MPC by allowing the humanoid robot to more effectively maintain balance and stability during walking, running, ju...
Balance control on uneven terrain
Uneven terrain can unpredictably and shift the robot’s center of mass and instantaneously re-distribute the feet forces. Thus, it is challenging balance whil...
Centroidal dynamics approximation
Centroidal dynamics approximation can significantly reduce the complexity of the model predictive control, enabling efficient computations for real-time pred...
Emerging data-driven approaches
Data-driven approaches have become increasingly relevant in robotic control as they have achieved unprecedented success in various applications, such as slid...
Floating-base state estimation
Accurately estimating the floating-base state is critical for humanoid robots as they move freely in space, but challenges such as noisy sensory data and tig...
High-order differential dynamics
Accurate dynamics model is essential for predicting future behavior and optimizing control actions over a finite time horizon.
Whole-body model predictive control
Whole-body Model Predictive Control (MPC) can enable humanoid robots to perform complex, dynamic tasks such as walking, running, jumping, and manipulation. ...
Control design on open-source platforms
Walking within limits: Implications of Hardware in Humanoid Locomotion with Centroidal Control.
Transfer human locomotion skills
Transferring human locomotion skills to humanoid robots is a multidisciplinary research field that combines biomechanics, machine learning, and neuroscience....