***Employment contracts can be signed in Hong Kong, but you must accept full-time or most of the time to work in Shenzhen!!!!

Duties:

1. Responsible for the research, development and debugging of biped humanoid robots/quadruped robots.

2. Responsible for building a foot robot simulation platform, designing the core motion simulation scene of the robot.

3. Promote the landing application of cutting-edge algorithms on the foot-type robot platform.

4. Construct and write a robot control-related SDK for internal colleagues to use.

5. Assist in the product design and planning, give design advice on mechanical, electronic and other hardware.

6. Develop deep reinforcement learning algorithms for multi-degree-of-freedom robots, bipedal robots, and quadrupedal robots for decision planning.

7. Responsible for model development, debugging, and actual robot verification of deep reinforcement learning algorithms.

Qualifications:

1. Computer science, artificial intelligence, robotics related master's degree or above, and with at least one year of research experience in the relevant direction.

2. Have research and development experience in foot-type robots, familiar with robot kinematics, dynamics, and common motion control algorithms.

3. Familiar with at least one mainstream simulation environment or framework, such as IsaacGym/Sim, PyBullet, MuJoCo, etc. Have real machine landing experience in solving reinforcement learning Sim-to-Real.

4. Have good computer programming skills (Python or C++ language), familiar with Linux, ROS operating system, good code conventions.

5. Familiar with the latest advances in learning-based control, and understand the commonly used deep reinforcement learning algorithms (PPO, SAC, DQN, DDPG, A3C, etc.).

6. Familiar with C++、Python programming language, familiar with mainstream deep learning frameworks such as pytorch/tensorflow, and understand frameworks such as ROS.

7. Have experience in deep reinforcement learning research projects applied to robots.