aifollow.news
Back IT之家 科技新闻
IT之家 科技新闻· · Original publication time AI score47

Hitch Open releases open-source physical AI platform; robot table-tennis rally reaches 142 shotsMachine translation

Hitch Open 开源发布物理智能平台,机器人乒乓对打达到 142 拍

AI-assisted summary

Hitch Open released its open-source physical AI competition platform globally at the China International Import Expo, focusing on perception, decision-making and action in real-world environments. A team led by Professor Gao Yue of Shanghai Jiao Tong University and Shanghai Innovation Institute worked with the platform on an AI-controlled robot table-tennis rally that reached 142 shots; the report says it set a world record for the event.

Article · AI translation

IT Home reported on November 8 that, according to Lingang Group, the ninth China International Import Expo hosted the global debut of a physical intelligence platform on November 7.

Hitch Open, a global open-source physical intelligence platform from Silicon Valley, California, whose academic foundations originate at the University of California, Berkeley, held a global launch event in the expo’s Lingang exhibition area and officially released its physical intelligence platform as open source worldwide.

IT Home learned that a challenge featuring robots playing table tennis against each other using autonomous AI decision-making was completed at the event. A team led by Professor Gao Yue of Shanghai Jiao Tong University–Shanghai Innovation Institute, together with the Hitch Open platform, achieved a new mark of 142 consecutive shots, breaking the world record for the event.

According to the organizers, Hitch Open is a competitive physical intelligence platform designed for real-world physical environments. Through demanding competitions in real-world settings, it aims to tackle the core challenges of enabling artificial intelligence to perceive, decide and act in the complex physical world, and to advance autonomous driving, embodied intelligence and robotics.

Playing table tennis against each other is an “ultimate test” for physical intelligence. Table tennis balls move quickly and their spin varies in complex ways, leaving robots a response window of only around 100 milliseconds and requiring them to assess the situation in real time within a closed loop of “perception–decision–action.” Each additional rally represents a simultaneous leap in the robots’ perceptual accuracy, decision-making speed and motion control capabilities.

Found an error? Send a correction