MIT Cheetah 3: Design and Control of a Robust, Dynamic Quadruped Robot

By Gerardo Bledt, Matthew J. Powell, Benjamin Katz, Jared Di Carlo, Patrick M. Wensing, Sangbae Kim. IEEE/RJS International Conference on Intelligent RObots and Systems, 2018

762 citations recorded by the source corpus.

This paper introduces a new robust, dynamic quadruped, the MIT Cheetah 3. Like its predecessor, the Cheetah 3 exploits tailored mechanical design to enable simple control strategies for dynamic locomotion and features high-bandwidth proprioceptive actuators to manage physical interaction with the environment. A new leg design is presented that includes proprioceptive actuation on the abduction/adduction degrees of freedom in addition to an expanded range of motion on the hips and knees. To make full use of these new capabilities, general balance and locomotion controllers for Cheetah 3 are presented. These controllers are embedded into a modular control architecture that allows the robot to handle unexpected terrain disturbances through reactive gait modification and without the need for external sensors or prior environment knowledge. The efficiency of the robot is demonstrated by a low Cost of Transport (CoT) over multiple gaits at moderate speeds, with the lowest CoT of 0.45 found during trotting. Experiments showcase the ability to blindly climb up stairs as a result of the full system integration. These results collectively represent a promising step toward a platform capable of generalized dynamic legged locomotion.

Subfields: Biomedical Engineering

View Paper: https://www.semanticscholar.org/paper/91a19620e40c3bad71b5c52eff40d86eaed32d29

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

RMA: Rapid Motor Adaptation for Legged Robots (2021, background): “The input channel number, output channel number, kernel size, and stride of each layer are [32, 32, 8, 4], [32, 32, 5, 1], [32, 32, 5, 1].”
Legged Locomotion in Challenging Terrains using Egocentric Vision (2022, neutral): “233 Several classical works use model based techniques, or define heuristic reactive controllers to 234 achieve the task of walking [16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28].”
Dynamic Locomotion in the MIT Cheetah 3 Through Convex Model-Predictive Control (2018, methodology): “The MIT Cheetah 3, shown in ﬁgure 1, is a 45 kg quadruped robot built by the MIT Biomimetic Robotics Lab [26].”
Rapid locomotion via reinforcement learning (2022, methodology): “Impressive progress has been made in this direction [5, 6, 9, 10, 11, 12, 15, 21].”
Learning to Walk in the Real World with Minimal Human Effort (2020, background): “For instance, researchers have demonstrated agile locomotion with quadrupedal robots using a state-machine [11], impulse scaling [42], and convex model predictive control [34].”
Mini Cheetah: A Platform for Pushing the Limits of Dynamic Quadruped Control (2019, methodology): “The mechanical design and actuation approach taken with the Mini Cheetah largely overlaps with its larger relative, the MIT Cheetah 3 [5].”
Learning Agile Robotic Locomotion Skills by Imitating Animals (2020, background): “The development of controllers for legged locomotion has been an enduring subject of interest in robotics, with a large body of work proposing a variety of control strategies for legged systems [37, 49, 54, 20, 18, 64, 8, 3].”
Learning to Walk via Deep Reinforcement Learning (2018, methodology): “nents such as state estimation, contact scheduling, trajectory optimization, foot placement planning, model-predictive control, and operational space control [1, 5, 13, 23].”
Data Efficient Reinforcement Learning for Legged Robots (2019, background): “Robust and agile locomotion of legged robots based on classical control stacks typically requires accurate dynamics models, human expertise, and tedious manual tuning [1, 2, 3, 4].”
Extreme Parkour with Legged Robots (2023, neutral): “Classical approaches for locomotion use model-based control to define walking controllers [15]– [19] and combine them with elevation maps constructed by fusing point cloud and odometry data [20]–[29].”