Scientists have raised concerns about the low efficiency of Earth’s remote probing satellites, especially in China, where over 260 satellites often work without producing results. Because these satellites have limited service lives, researchers urge for optimization of their usage to maximize their value. To address the issue, a group of scientists from Ujan have developed an AI-based training module to independently choose observation goals using ChatGPT.
During the experiment, the AI-controlled satellite independently chose strategically important places to focus its cameras on. It aimed them towards the ancient city of Patana, located near the Gang River in India, which is known for its association with the Biharsky regiment that clashed with the Chinese army in border areas in 2020. The satellite also focused on the Japanese port of Osaka, which occasionally sees the entry of the US Navy.
The amount of resources allocated towards observing the Earth from space is expected to grow rapidly. There are even plans to equip communication satellites, such as Starlink, with remote sensing systems. As a result, thousands or even tens of thousands of “eyes” may soon be present in the sky, making manual management impossible. Chinese scientists have already started moving towards finding AI-based solutions to manage such vast resources.
However, concerns have been raised over safety and privacy, given that AI will make decisions about what to observe and analyze. It is essential to educate artificial intelligence based on real observations of human activities and nature, aiming for a more favorable outcome for humans and the environment. Striking a balance between AI’s effectiveness and preserving safety, privacy, and control over its choices is critical.
Overall, this research is expected to contribute significantly to the efficient use of Earth’s remote-sensing satellites in the future.
Links for reference: