This year, the National-Local Co-built Humanoid Robot Innovation Center, located in the Beijing Economic-Technological Development Area (Beijing E-Town), has been focusing on three aspects of technological innovation, product application and ecosystem construction, to accelerate the development of humanoid robots.

Currently, the center is using data sets to break down and execute long-distance complex tasks for humanoid robots so they can be applied in more scenarios.

Since the release of the RoboMIND data set in December, the center has been continuously improving the data set, investing in larger spaces, more scenarios and more personnel for data collection and annotation. To date, 100,000 high-quality intelligent data points have been open-sourced, allowing more research institutions and enterprises to use this data set to better train humanoid robots.

"The research and development tuning we are currently conducting is also the preparation for 'Tiangong' to participate in the humanoid robot marathon to be held in E-Town in April this year," said Wei Jiaxing, a staff member of the center.

Tiangong, a robot developed by the center, has recently been re-adjusted to reduce its weight, improve leg flexibility, increase battery capacity and optimize core components such as joints for load-bearing and shock absorption.

Furthermore, by equipping Tiangong with a perception system, it now has improved autonomous adaptability in complex environments and also possesses stronger anti-interference capabilities.

"By participating in sports competitions, we hope to better educate the public and let everyone know how far robots have developed. This will also help promote humanoid robots to truly enter every household in the future," said Wei.

Next, the center will collaborate with global industry-academia-research partners to continuously innovate in embodied intelligence core technologies and explore the application value of multiple scenarios and tasks in industrial manufacturing, logistics sorting, elderly care services and specialized operations. Meanwhile, the center will also promote the continuous open-source of embodied intelligent data sets to accelerate industry development.