Feedback Control of a Cassie Bipedal Robot: Walking, Standing, and Riding a Segway

By Yukai Gong, Ross Hartley, Xingye Da, Ayonga Hereid, Omar Harib, Jiunn-Kai Huang, Jessy W. Grizzle. 2018

220 citations recorded by the source corpus.

The Cassie bipedal robot designed by Agility Robotics is providing academics with a common platform for sharing and comparing algorithms for locomotion, perception, and navigation. This paper focuses on feedback control for standing and walking using the methods of virtual constraints and gait libraries. The designed controller was implemented six weeks after the robot arrived at the University of Michigan and allowed it to stand in place as well as walk over sidewalks, grass, snow, sand, and burning brush. The controller for standing also enables the robot to ride a Segway. Software supporting the work in this paper is available on GitHub.

Subfields: Artificial Intelligence, Biomedical Engineering, Mechanical Engineering

How researchers see this paper

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

In Artificial Intelligence

1. Sergey Levine
2. Xue Bin Peng
3. Pieter Abbeel
4. Michiel van de Panne
5. Chelsea Finn
6. Glen Berseth
7. Ruslan Salakhutdinov
8. Jitendra Malik

In Biomedical Engineering

1. Sergey Levine
2. Pieter Abbeel
3. Xue Bin Peng
4. Michiel van de Panne
5. Aaron D. Ames
6. Koushil Sreenath
7. Glen Berseth
8. Jake Reher

In Mechanical Engineering

1. Pieter Abbeel
2. Sergey Levine
3. Xue Bin Peng
4. Zhaoming Xie
5. Jeremy Dao

Citations in the corpus

• Learning quadrupedal locomotion over challenging terrain (2020, neutral): “Many rely on elaborate state machines that coordinate the execution of motion primitives and reflex controllers [1–5].”
• Dynamic Walking with Compliance on a Cassie Bipedal Robot (2019, methodology): “Therefore, certain levels of approximation have been suggested: a gait library method [8] which ignores the compliant dynamics in the robot has demonstrated robust walking behaviors with the addition of several ad-hoc components; a machine learning method [25] has also been used to train control pol”
• Reinforcement Learning for Robust Parameterized Locomotion Control of Bipedal Robots (2021, methodology): “Model-based methods require careful system identification to bridge the reality gap [16], [17].”
• Learning Locomotion Skills for Cassie: Iterative Design and Sim-to-Real (2019, methodology): “To utilize the full dynamics of the robots, trajectory optimization [19] is used to generate trajectories, and tracking controllers based on quadratic programming [20] or feedback linearization [3] can be designed along them.”