On January 8, the Siasun Robot&Automation embodied algorithm team announced the launch of the EnerVerse architecture, which guides robots in completing complex tasks while generating future embodied spaces through an autoregressive Diffusion model. Unlike existing methods that simply apply video generation models, EnerVerse introduces a Sparse Memory mechanism and Free Anchor View (FAV), enhancing 4D generation capabilities. According to Siasun Robot&Automation, experimental results indicate that EnerVerse not only possesses exceptional future space generation capabilities but also achieves state-of-the-art performance in robot motion planning tasks.