China-Humanoid Robot Games/Training

Robotics firm trains for scenario challenges for upcoming World Humanoid Games

  • English
DOWNLOAD
  • ID : 8494893
  • Dateline : Recent
  • Location : China
  • Category : Tech
  • Duration : 1'52
  • Audio Language : Chinese/Nats/Part Mute
  • Source : China Central Television (CCTV)
  • Restrictions : No access Chinese mainland
  • Published : 2026-08-18 16:39
  • Last Modified : 2026-08-18 16:43:35
  • Version : 1

China-Humanoid Robot Games/Training

Robotics firm trains for scenario challenges for upcoming World Humanoid Games

Dateline : Recent

Location : China

Duration : 1'52

  • English


Beijing, China - Recent (CCTV - No access Chinese mainland)
1. Various of humanoid robot arranging books
2. Various engineer controlling robot
3. SOUNDBITE (Chinese) Zhou Wei, head, scenario-based events team, Realman Robotics (ending with shots 4-6):
"First, we need to see clearly. The display on the operation interface needs to be as close to reality as possible. We have done a great deal of work to streamline the data link and optimize low-latency data transmission."
4. Various of robotic arms
5. Various of computer screen displaying robot arranging books
6. Various of engineers controlling robots to perform tasks
7. Various of robotic arms
8. SOUNDBITE (Chinese) Zhou Wei, head, scenario-based events team, Realman Robotics (starting with shot 7/ending with shots 9-12):
"Our humanoid arm is relatively smaller and slimmer, more resembling a human arm, but it can actually bear much greater loads than other joints and arms of the same size. For this competition, we have optimized 360-degree full-field perception, enabling the robot to see all perspectives in the space across all terrains. We have also adjusted the robot's mobility, load capacity, and end effectors. We need to use competitions with real scenarios to truly train our embodied intelligent robot to become a human partner or a collaborative partner for humans."
9. Various of sensors, cameras on robot
10. Screen showing data
11. Robotic component
12. Various of robots moving, performing tasks


A robotic data training ground has been set up to simulate multiple real-world scenarios and conduct targeted training ahead of the upcoming World Humanoid Robot Games scheduled to open in Beijing on August 22.

Jointly organized by the Beijing Municipal Government and China Media Group, the much-anticipated Games have drawn 666 teams and 2,056 robots from 16 countries. The five-day event includes 30 competitive programs spanning ball games and combat categories. There are also 21 scenario-based challenges covering industrial assembly and home services, which kicked off on Sunday.

Realman Robotics has installed the data training ground and prepared for the competition. A library scene has been set up, where robots are tasked with arranging books. The robots must select books of the correct color and precisely insert them onto the shelves with the spines facing outward, without knocking over any other books. The biggest challenge in this process is the precision with which operators control the robots.

"First, we need to see clearly. The display on the operation interface needs to be as close to reality as possible. We have done a great deal of work to streamline the data link and optimize low-latency data transmission," said Zhou Wei, head of the scenario-based events team at Realman Robotics.

To prepare for the scenario-based events at the competition, the company has set up training grounds for more than 10 simulated competition scenarios, including hotel bedrooms, pharmacies and firefighting scenes.

The upcoming household service event covers tasks such as sorting out clothes, tidying up, and daily cleaning. These seemingly ordinary household chores represent far more complex technological challenges for robots.

"Our humanoid arm is relatively smaller and slimmer, more resembling a human arm, but it can actually bear much greater loads than other joints and arms of the same size. For this competition, we have optimized 360-degree full-field perception, enabling the robot to see all perspectives in the space across all terrains. We have also adjusted the robot's mobility, load capacity, and end effectors. We need to use competitions with real scenarios to truly train our embodied intelligent robot to become a human partner or a collaborative partner for humans," Zhou said.

ID : 8494893

Published : 2026-08-18 16:39

Last Modified : 2026-08-18 16:43:35

Source : China Central Television (CCTV)

Restrictions : No access Chinese mainland

More



Login
Username
Password
code
Sign In
OK