Robotic hands work on a new energy vehicle (NEV) production line of NIO Second Advanced Manufacturing Base on July 1, 2025 in Hefei, Anhui Province of China. The NIO Second Advanced Manufacturing Base is a digital intelligent factory with workshops such as stamping, car body, painting and final assembly. Photo Credit: China News Service/Getty Images
Technology USAAI Teaches Robots to Work Together Seamlessly
A new breakthrough by Google DeepMind could change the way factories operate. The team has developed RoboBallet, an AI system that enables manufacturing robots to coordinate their movements autonomously — performing complex tasks without colliding or wasting time.
“Solving all three of these problems combined is what we tackled in our work,” said Matthew Lai, research engineer at Google DeepMind, referring to the challenges of task allocation, scheduling, and motion planning.
Trained using graph neural networks, RoboBallet can choreograph up to eight robots performing 40 tasks with remarkable speed and precision. It generates complete production plans in seconds, compared to the hundreds of hours traditional programming requires. In trials, it achieved near-human efficiency and even adapted automatically when one robot malfunctioned. The team hopes this system will soon make manufacturing faster, safer, and far more flexible — turning factory floors into perfectly synchronised performances.
