Robotics dogs, Tencent Robotics X Lab, Tencent Holdings Limited, Shenzhen, Guangdong, China

[Airicist]

Developer - Tencent Robotics X Lab (https://pr.ai/showthread.php?t=11077)

[Airicist]

https://youtu.be/BzSgdnI2okk

Meet Jamoca, the robot dog

Nov 30, 2020


The Kungfu Dog! Meet Jamoca, a robot dog newly released by Tencent's Robotics X Lab. It shows us its impressive skills by walking on some plum blossom poles. Take a look!

"Tencent robotics x lab quadruped mobile robot Jamoca debut (https://equalocean.com/briefing/20201120230004679)"

November 20, 2020

[Airicist]

https://youtu.be/BXJ_eBQmtfE

Tencent released the first fully self-developed quadruped robot

Mar 2, 2021

"Tencent unveils Max, its first in-house developed quadruped robot (https://cntechpost.com/2021/03/02/tencent-unveils-max-its-first-in-house-developed-quadruped-robot)"

by Tom Kang
March 2, 2021

[Airicist]

https://youtu.be/g_0dtaTxG2M

Run like a dog: learning based whole-body control framework for quadruped gait style transfer

Jul 2, 2021


Video attachment for the contributed paper accepted by 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems.

Title: Run Like a Dog: Learning Based Whole-Body Control Framework for Quadruped Gait Style Transfer

Authors: Fulong Yin, Annan Tang, Liangwei Xu, Yue Cao,
Yu Zheng, Xiangyu Chen, Zhengyou Zhang

Abstract: In this paper, a learning-based whole-body locomotion controller is proposed, which enables quadruped robots to perform running in the style of real animals. We use alow-level controller based on multi-rigid body dynamics to calculate desired torques for each joint, while the high-level neural network policy planning the expected gait and foothold. The policy is trained with reinforcement learning so that the robot can track a variety of trajectories according to the gait patterns recorded from real-world dogs. We transfer the walking and running gait style to quadrupeds in simulation, involving pace, trot, high-speed gallop, and natural transitions. The performance is evaluated by the synchronization rate of the contact state between the policy result and the recorded sequence. In the experiments, the robot runs steadily at a speed of 2 m/s and showcases a notable synchronization rate of about 80%. Without prior knowledge, the policy demonstrates a realistic foothold distribution that covers the central area of the torso, which is prevalent in running animals.

Acknowledgment: This work is supported by Tencent Robotics X.