ANYmal - a highly mobile and dynamic quadrupedal robot

By Marco Hutter, Christian Gehring, Dominic Jud, Andreas Lauber, Dario Bellicoso, Vassilios Tsounis, Jemin Hwangbo, K. Bodie, Péter Fankhauser, Michael Bloesch and 4 more. 2016

991 citations recorded by the source corpus.

Subfields: Biomedical Engineering

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

Citations in the corpus

Learning quadrupedal locomotion over challenging terrain (2020, methodology): “The controller was used to drive two generations of ANYmal quadrupeds [20] in a variety of conditions that are beyond the reach of prior published work in legged robotics.”
Learning agile and dynamic motor skills for legged robots (2019, methodology): “Most legged systems are driven by hydraulic actuators [51] or electric motors with gears [3], and some even include dedicated mechanical compliance [5, 52].”
Probabilistic foot contact estimation by fusing information from dynamics and differential/forward kinematics (2016, methodology): “In case of dynamic robots like StarlETH [1] and its decedent ANYmal [2], adding force sensors at the feet is undesirable due to the following three reasons.”
RMA: Rapid Motor Adaptation for Legged Robots (2021, methodology): “The ANYmal robot [20] locomotes by optimizing a parameterized controller and planning based on an inverted pendulum model [15].”
Dynamic Locomotion Through Online Nonlinear Motion Optimization for Quadrupedal Robots (2018, methodology): “Over the past few years, ANYmal [8] (see Fig.”
Robust Recovery Controller for a Quadrupedal Robot using Deep Reinforcement Learning (2019, methodology): “The system was tested on the quadrupedal robot ANYmal [10] in more than 100 trials with a success rate higher than 97 %.”
Mini Cheetah: A Platform for Pushing the Limits of Dynamic Quadruped Control (2019, methodology): “A similar approach was used in the quadruped ANYmal, with the series-elastic ANYdrive actuator [6].”
Where Should I Walk? Predicting Terrain Properties From Images Via Self-Supervised Learning (2019, methodology): “We deploy and test the proposed system on the ANYmal quadruped [6].”
MIT Cheetah 3: Design and Control of a Robust, Dynamic Quadruped Robot (2018, background): “By comparison, the series elastic actuators used in ANYmal, which measure the deflection of a series spring to control output torque, have far greater torque accuracy, but much lower peak torque, speed, and torque bandwidth, especially at high torques [2].”
Extreme Parkour with Legged Robots (2023, neutral): “…design principle for general robot base motion acquisition, together with an automatic terrain curriculum for overcoming exploration burden in RL. • Our results set a new landmark in learning-driven park-our with high jumps that are 2x the height of the robot, long jumps that are 2x the…”
Benchmarking Classic and Learned Navigation in Complex 3D Environments (2019, background): “Methods of this type have been successfully deployed on a variety of mobile robotic platforms [41, 4, 20].”
Dynamic Locomotion on Slippery Ground (2019, methodology): “To verify our approach, we have conducted several experiments on ANYmal [19], a fully torque-controllable quadrupedal robot.”
Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learning (2021, background): “The reward function is as follows: reaching cost ANYmal Locomotion For the included ﬂat-terrain environment, observations are detailed in Table 8 and the reward function is as follows: For rough terrain locomotion with sim-to-real, we extend the observations with 140 terrain heights around the robot”
Whole-Body Nonlinear Model Predictive Control Through Contacts for Quadrupeds (2017, methodology): “We show that the approach transfers between platforms, and apply the same framework to the two quadrupeds HyQ and ANYmal (Figure 1).”
Per-Contact Iteration Method for Solving Contact Dynamics (2018, methodology): “We test the following four scenarios: • Trotting: The model of ANYmal runs forward at 1.5 m/s with its legs moving in diagonal pairs.”
Gait and Trajectory Optimization for Legged Systems Through Phase-Based End-Effector Parameterization (2018, methodology): “Secondly, we demonstrate a subset of motions in the realistic physics simulator Gazebo as well as on the quadruped robot ANYmal [34].”
Legged Locomotion in Challenging Terrains using Egocentric Vision (2022, methodology): “Several classical works use model based techniques, or define heuristic reactive controllers to achieve the task of walking [19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31].”