Teaching robots to jump and maintain their balance is easier said than done. With reinforcement learning, researchers are pulling off impressive feats. This video shows the robot Olympus performing vertical jumps of 1m and horizontal jumps of 1.25m. The below video shows it in action:
More like this ➡️ here
Towards Quadrupedal Jumping and Walking for Dynamic Locomotion using Reinforcement Learning
The researchers also came up with a walking policy that combines with jumping policies for dynamic locomotion. This robot can perform high jumps with cm accuracy. This method can be used to learn omnidirectional jumping.
[HT]
[HT]
*Our articles may contain aff links. As an Amazon Associate we earn from qualifying purchases. Please read our disclaimer on how we fund this site.
