FILE: In Space - June 4, 2024 (CCTV- No access Chinese mainland)
1. Photo of Chang'e-6 lunar probe's lander, ascender on surface of moon's far side with Chinese national flag

China - Recent (CCTV - No access Chinese mainland)
2. Various of autonomous intelligent micro-robot for moon surface at its testing ground
3. SOUNDBITE (Chinese) Xing Yan, staff member, China Aerospace Science and Technology Corporation (starting with shot 2/ending with shot 4):
"Though the droid is very tiny, it is very functional. It is capable of autonomous separation, autonomous movement and lunar imagining. After reaching the surface, it can detach from the lander by itself, and move to a relatively suitable location for taking pictures, and it was able to intelligently choose the best angle and composition, leaving some precious memories for the Chang'e-6 mission."
4. Various of mini rover driving over obstacle; staff of China Aerospace Science and Technology Corporation (CASC)
5. Various of CASC staff working, studying mini rover
6. SOUNDBITE (Chinese) Xing Yan, staff member, China Aerospace Science and Technology Corporation (starting with shot 5/ending with shot 7):
"We are enhancing its autonomy with AI technology, and we're still using neural networks. We give full play to neural networks' ability to learn from human experience, and it can achieve many things that cannot be achieved with traditional methods, such as adjusting the angle of the photo. This also verifies the feasibility of this AI technology based on neural networks under the constraints of limited computing resources on the lunar surface."
7. Various of staff working on laptop; rover wheels
8. Various of rover riding on testing ground; staff working; wheels
9. SOUNDBITE (Chinese) Huang Huang, staff member of China Aerospace Science and Technology Corporation (starting with shot 8/ending with shot 10):
"We are more concerned about the position of the main body's position and the angle of some sunlight. The reflection may be too strong in some angles, which is not our ideal situation."

FILE: In Space - June 4, 2024 (CCTV- No access Chinese mainland)
10. Photo of Chang'e-6 lunar probe's lander, ascender on surface of moon's far side

China - Recent (CCTV - No access Chinese mainland)
11. Huang working with colleague
12. Animation of Chang'e-6 lunar probe's lander, ascender
13. Various of Huang working with colleagues, rover
14. SOUNDBITE (Chinese) Huang Huang, staff member of China Aerospace Science and Technology Corporation (starting with shot 13):
"Intelligence algorithm is to simulate people taking photos of an object. It does not need to run precise calculations, but to make independent judgments on whether the photos taken are good or bad, make some decisions or adjustments, and finally achieve the best results we want."
15. Various of staff working, rover riding; Huang working
16. SOUNDBITE (Chinese) Huang Huang, staff member of China Aerospace Science and Technology Corporation (starting with shot 15/partially overlaid with shot 17/ending with shot 18):
"I was very impressed by what happened at that time - in order to solve such problems, we bought floorlamps, camping lights, stage lamps, and many other kinds of lights online. We tried our best to make it as knowledgeable as possible, to construct these physical samples, to do some feature modeling of the environment, and in such an environment, we let the rover learn and finally realize the intelligent decision support when taking pictures."
++SHOT OVERLAYING SOUNDBITE++
17. Mini rover
++SHOT OVERLAYING SOUNDBITE++
18. Various of mini rover on testing ground, Huang working with colleagues

The photo of Chang'e-6 lunar probe's lander and ascender on the surface of the far side of moon that marked a historic moment of China's space exploration was taken by a tiny five-kilo mini robot rover powered by autonomous piloting and artificial intelligence (AI) camera technology.

The autonomous and intelligent micro-robot for moon surface is highly functional despite its small size as it is equipped with significantly enhanced autonomous intelligence and highly integrated, lightweight hardware.

The mini rover detached from the Chang'e-6 lander after the sampling mission and moved to an optimal position. Using onboard artificial intelligence and neural networks, it composed and captured the third-person view without human input.

"Though the droid is very tiny, it is very functional. It is capable of autonomous separation, autonomous movement and lunar imagining. After reaching the surface, it can detach from the lander by itself, and move to a relatively suitable location for taking pictures, and it was able to intelligently choose the best angle and composition, leaving some precious memories for the Chang'e-6 mission," said Xing Yan, staff member of China Aerospace Science and Technology Corporation (CASC).

As a key highlight of the rover, the autonomous intelligence, which is similar to AI technology, has the ability to think and judge more like a person, developers at the CASC said.

Neural networks give the robot abilities to learn from experiences and make adjustments like changing photographic angles, Xing said.

"We are enhancing its autonomy with AI technology, and we're still using neural networks. We give full play to neural networks' ability to learn from human experience, and it can achieve many things that cannot be achieved with traditional methods, such as adjusting the angle of the photo. This also verifies the feasibility of this AI technology based on neural networks under the constraints of limited computing resources on the lunar surface," said Xing.

"We are more concerned about the position of the main body's position and the angle of some sunlight. The reflection may be too strong in some angles, which is not our ideal situation," said Huang Huang, another staff member of CASC.

Previous photo missions on the Moon required pre-calculation on the relative distance between the subject and the camera, the orientation and the best time to take the picture. But this time with the use of autonomous intelligent photography, it all relies on the robot to judge when and how to make the best shot.

"Intelligence algorithm is to simulate people taking photos of an object. It does not need to run precise calculations, but to make independent judgments on whether the photos taken are good or bad, make some decisions or adjustments, and finally achieve the best results we want," Huang said.

To simulate lighting on the moon's dark side with parallel light from the sun being the only optical source, CAST staff trained the bot using different lamps during development.

"I was very impressed by what happened at that time - in order to solve such problems, we bought floorlamps, camping lights, stage lamps, and many other kinds of lights online. We tried our best to make it as knowledgeable as possible, to construct these physical samples, to do some feature modeling of the environment, and in such an environment, we let the rover learn and finally realize the intelligent decision support when taking pictures," she said.

CAST developers hailed that the successful lunar sampling by Chang'e-6 marking a groundbreaking achievement in the history of human lunar exploration, adding that the precious images recorded by the micro-robot will become also timeless classics.

More importantly, this innovative practice signifies a solid step toward autonomous intelligence in deep space exploration, promising to have a positive impact on future lunar scientific research, they said.