Tesla showed off the latest progress of the Optimus robot on the Twitter (X) platform, with videos showing visual self-calibration, color sorting tasks, balancing on one foot, and more. From a technical point of view, Tesla has demonstrated the possibility of robot End-to-end technology route, relying on visual self-calibration to generate real-world data, training the end-to-end network through Dojo, directly generating robot control instructions from images and executing, and achieving color sorting tasks. We believe that this solution is similar to its FSD model, and the successful experience of FSD and Dojo has the potential to accelerate the speed of robot iteration from single task to multi-task expansion. It is recommended to pay attention to Tesla robot-related benefit targets and companies that are expected to benefit from Tesla's lead in robot end-to-end network and robot attitude control.