Recently, Tsinghua University, in collaboration with Xinqiong and Zhiyuan, proposed a group experiential learning paradigm and completed large-scale verification in a cluster of 96 real robots. CEL allows robots not only to learn by human demonstration, but also to learn from the successful experiences of other robots in the group. By sharing and accumulating experience, robot groups can jointly improve their capabilities in continuous practice, providing a new technical path for promoting autonomous evolution and large-scale application of robots.

Recently, Tsinghua University, in collaboration with Xinqiong and Zhiyuan, proposed a group experiential learning paradigm and completed large-scale verification in a cluster of 96 real robots. CEL allows robots not only to learn by human demonstration, but also to learn from the successful experiences of other robots in the group. By sharing and accumulating experience, robot groups can jointly improve their capabilities in continuous practice, providing a new technical path for promoting autonomous evolution and large-scale application of robots.