BY MARIA BURKE | 21 DECEMBER 2023
Chinese researchers have developed an AI robot that can make oxygen from ores found in Martian meteorites. They say similar technology could one day make human missions to Mars easier and more cost-effective.
Potential future manned missions to Mars will require oxygen as it is essential to human activity on the planet, being used in rocket propellants and life-support systems.
One of the ways to make missions simpler and cheaper would be to use resources already present on the planet to create oxygen, rather than transport materials from Earth. Recent evidence of water on Mars and analysis of Martian meteorites suggest it should be possible to make oxygen catalysts.
There are two major technical challenges. First, any system needs to be unmanned and self-directing. Second, it needs to identify the best catalyst ingredients, given knowledge of the existing elements in Martian ores. Designing a catalyst from a given list of elements is a daunting task using conventional ‘trial-and-error’.
Instead, Jun Jiang of the University of Science and Technology of China and colleagues have designed a ‘robotic chemist’ based on self-learning AI algorithms.
They asked it to analyse five categories of meteorites confirmed to exist on Mars. It was able to identify chemicals in the meteorites that could react together to produce oxygen. The robot then made catalysts from these compounds before testing their oxygen production performance. This process was repeated until it found the best formula for a Martian oxygen catalyst from more than three million possible combinations – something which the researchers estimate could have taken humans 2000 years (Nature Synthesis, doi: 10.1038/s44160-023-00424-1).
The researchers say the study demonstrates that the robot can perform all the necessary experimental steps, including raw material analysis, pretreatment, synthesis, characterisation and performance testing with high precision, without human intervention and under simulated Martian conditions. ‘Particularly powerful’ is its ability to combine experimental and computational data during the synthesis process.
Visit Blog
