Learning Active Spine Behaviors for Dynamic and Efficient Locomotion in Quadruped Robots

05/15/2019
by   Shounak Bhattacharya, et al.
0

In this work, we provide a simulation framework to perform systematic studies on the effects of spinal joint compliance and actuation on bounding performance of a 16-DOF quadruped spined robot Stoch 2. Fast quadrupedal locomotion with active spine is an extremely hard problem, and involves a complex coordination between the various degrees of freedom. Therefore, past attempts at addressing this problem have not seen much success. Deep-Reinforcement Learning seems to be a promising approach, after its recent success in a variety of robot platforms, and the goal of this paper is to use this approach to realize the aforementioned behaviors. With this learning framework, the robot reached a bounding speed of 2.1 m/s with a maximum Froude number of 2. Simulation results also show that use of active spine, indeed, increased the stride length, improved the cost of transport, and also reduced the natural frequency to more realistic values.

READ FULL TEXT

page 1

page 2

page 4

page 5

research
05/03/2023

Enhancing Efficiency of Quadrupedal Locomotion over Challenging Terrains with Extensible Feet

Recent advancements in legged locomotion research have made legged robot...
research
07/07/2021

Quadruped Locomotion on Non-Rigid Terrain using Reinforcement Learning

Legged robots need to be capable of walking on diverse terrain condition...
research
07/23/2021

Bio-inspired Rhythmic Locomotion in a Six-Legged Robot

Developing a framework for the locomotion of a six-legged robot or a hex...
research
01/03/2019

Design, Development and Experimental Realization of a Quadrupedal Research Platform: Stoch

In this paper, we present a complete description of the hardware design ...
research
03/20/2022

Hierarchical Reinforcement Learning of Locomotion Policies in Response to Approaching Objects: A Preliminary Study

Animals such as rabbits and birds can instantly generate locomotion beha...
research
11/11/2020

Learning Agile Locomotion Skills with a Mentor

Developing agile behaviors for legged robots remains a challenging probl...
research
01/26/2021

Online Body Schema Adaptation through Cost-Sensitive Active Learning

Humanoid robots have complex bodies and kinematic chains with several De...

Please sign up or login with your details

Forgot password? Click here to reset