Source: REUTERS/Denis Balibouse
[Saba Sports News] According to the latest news from Swiss media, a team led by ETH Zurich recently developed an artificial intelligence-driven badminton robot named “ANYmal-D” designed to play against humans. This robot uses reinforcement learning technology, a type of artificial intelligence that improves decision-making through repeated attempts. The robot interacts with its environment, tries different actions, and adjusts its behavior based on feedback rewards, gradually learning how to make optimal decisions in badminton matches. During tests against human players, ANYmal-D successfully returned balls at various speeds and angles, achieving up to 10 consecutive hits. It can handle up to 12.06 meters per second badminton speed and dynamically adjust its gait based on the flight time and distance of the ball. Although not yet mature, it has already been able to achieve up to 10 consecutive hits. I believe this demonstrates the potential of AI-driven legged robots to perform complex tasks, paving the way for the development of future autonomous intelligent systems, including humanoid robots.