Video Transcript
I've never said that life as a robot is all about work and not play.
Eth Zurich shows what any Mal-D robot will do in his leisure time, and its ability suggests that it may be the next opponent in a casual badminton game.
AI-enabled four-legged robots have recently proven that they can play autonomous badminton against human enemies.
According to the research team, robots can track the shuttlecock paths and trajectories using a control system driven by reinforcement learning. The bot can also use a dynamic arm to move across the court to hold the racket and return a shot.
However, sports robots are not the main goal of this technology. The aim is to develop a method to understand how to combine precision sensing with rapid systemic responses.
Researchers have said that replication of this type of movement with leg robots has been a challenge so far, as there are limitations to the controllers and hardware currently available. They say that the human eye is actually much better than a robotic camera when it comes to focus and stabilizing movement.
The Unified RL-based controller is behind Any-Mal-D's movement and uses onboard recognition autonomously. This is what they say is “visual errors due to motion,” bringing the device's capabilities closer to human capabilities.
Yuntao Ma, a researcher at the Robotics Lab in Eth Zurich and co-author of the study, told an interesting engineer:
And while for now you might be able to beat Any-Mal-D in badminton, don't be too comfortable. Limiting the ability to respond quickly enough to offensive shots is an issue that can be addressed in the future with additional sensing modalities or faster cameras.
