On October 30, Zhiyuan teamed up with Tsinghua University and Infinigence AI to unveil the Collective Experience Learning (CEL) paradigm. This paradigm draws on the real operational trajectories of robot peers as its source of experience, complemented by the FERT failure-guided experience retrieval fine-tuning algorithm and the FATS fault-tolerant asynchronous training system. Across five real-world tasks—flower arrangement, clothes folding, mobile phone charging, stool assembly, and mobile phone packaging—the policies trained with CEL consistently outperformed the average level of six human teleoperators, with a lead of up to 16.7 percentage points. In the mobile phone packaging task, the number of training robots scaled from 16 to 96, and the success rate climbed from 86.7% to 98.3%.