In recent years, there has been a lot of news in the field of humanoid robots. The eight immortals cross the sea and show their magic power. In this track with a history of more than 50 years, there are also stories from zero to one in half a year.

In March this year, the team Zhejiang Humanoid Robot Innovation Center, which has iterated for 18 years in the field of humanoid robots, released the universal humanoid robot "Leader 1 」.

The core members of the team are from Zhejiang University and the top domestic design team.
At the beginning of the design, this design was jointly selected with the Zhejiang University team from nearly 100 groups of design schemes, mainly based on lightweight, sense of service, and sense of technology as the basic design standards.

Throughout the design process, we have always adhered to our original intention: innovation-oriented, designing a better life, working closely with structure, research and development, and engineering, and finally presented the leader No. 1 in March this year.

"Pilot No. 1" adopts the style of the close integration of simulated learning and linear motion control. It takes into account the execution ability of the machine while learning the actions of the people. It opens up the whole link of the action mapping to the real object and can be generalized, endowing the robot with the ability of human learning.

At the same time, "Pilot 1" builds a model-free end-to-end servo control, does not rely on large models, does not rely on object models, can dynamically adapt to the environment operation, get rid of the dependence on cloud computing power, so that the local deployment of the algorithm is more convenient, some algorithms have been applied in the head industrial enterprises and have been commercially recognized.

In the collaborative control of the robot's arms and legs, the "pilot 1" also tries to use the combination of mechanism modeling and reinforcement learning to ensure the reliability of adaptation and execution to different environments, and realize the whole body operation control of the robot.