Design Principles for a Family of Direct-Drive Legged Robots

By Gavin D. Kenneally, Avik De, Daniel E. Koditschek. 2016

311 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Control and Systems 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

In Control and Systems Engineering

Citations in the corpus

Sim-to-Real: Learning Agile Locomotion For Quadruped Robots (2018, methodology): “We also design simple experiments to measure motor frictions [20].”
Soft Actor-Critic Algorithms and Applications (2018, methodology): “We use the Minitaur robot, a small-scale quadruped with eight direct-drive actuators (Kenneally et al., 2016).”
Mini Cheetah: A Platform for Pushing the Limits of Dynamic Quadruped Control (2019, methodology): “Both the predecessor to this robot, the Super Mini Cheetah [1], and Minitaur [2] feature highly backdriveable actuators, making them mechanically robust and allowing control of ground reaction forces through proprioception, but have only two degrees of freedom per leg.”
ANYmal - toward legged robots for harsh environments (2017, methodology): “Similar concepts were also adopted in theMinitaur project [16], a comparably small and robust quadrupedal device.”
Modular Hopping and Running via Parallel Composition (2017, methodology): “…The electromagnetic motors and the transmission at their outputs can very easily hamper system bandwidth in significant ways, as we described in Kenneally et al. (2016)—motors from “reputable” motor manufacturers such as Maxon are not, in fact, the best ones available based on physically…”
Data Efficient Reinforcement Learning for Legged Robots (2019, methodology): “We use the Minitaur robot from Ghost Robotics [25] as the hardware platform for our experiments.”
Learning to Walk in the Real World with Minimal Human Effort (2020, methodology): “In the first experiment, we test our system using a Minitaur, a quadruped robot (Figure 1) [43], to learn forward and backward walking in a 5× 2m training area with flat ground.”
Learning to Walk via Deep Reinforcement Learning (2018, methodology): “We demonstrate the framework by training a Minitaur robot [26] (Figure 1) to walk.”
Robotic vertical jumping agility via series-elastic power modulation (2016, background): “1, eaag2048 (2016) 6 December 2016 (19) and theMIT (Massachusetts Institute of Technology) Cheetah (20) have leveraged the fact that electromagnetic actuators scale favorably with size (21) and use rigid, torque-dense actuators.”