NASA officials say they hope to use artificial intelligence, such as machine learning, to analyze data collected by large telescopes to help search for extraterrestrial life and detect asteroids that pose a potential threat to Earth, SpaceNetwork reported. NASA has partnered with companies such as Intel, IBM and Google to develop advanced machine learning technologies. Every summer, NASA brings together technology and space innovators for an eight-week Frontier Development Laboratory (FDL) project.
In 2018, NASA Goddard Space Flight Center astrobiologist Jada Arnie and colleague Sean Domagal-Goldman led an FDL team that developed a machine learning technique that uses brain-like “neural networks” to analyze images of exoplanets. The chemical properties of the planet are identified according to the wavelength at which molecules in the planet’s atmosphere emit (or absorb) light. The way this technology processes information is the same way that neurons in the brain process and transmit information to each other. The results show that, using this neural network technology, the researchers were able to identify the abundance of various molecules in the atmosphere of the exoplanet WASP-12b, which is more accurate than conventional methods.
The researchers say that while their neural network technology is still in its infancy, it could one day be used to study data collected by future telescopes to help screen exoplanets that deserve further study. “The data we get in the future can be so vast that it’s hard to understand, and AI tools are expected to benefit us,” Arnie said. “
In addition, another team developed a machine learning program in 2017 that creates three-dimensional models of asteroids in just four days, including their size, shape and rotation speed. The researchers say such procedures are particularly important in detecting asteroids that could threaten the planet from Earth.
NASA officials say the agency’s spacecraft can provide about 2GB of data every 15 seconds, but “limited to manpower, time, and resources, we can only analyze a small amount of data, and artificial intelligence can help us in this area.” “
The researchers also suggest that integrating AI technology into future spacecraft will allow them to make real-time scientific decisions, saving spacecraft time for communication to the ground. Of course, Arnie also stressed that the AI tools will not replace humans any time soon, as the results still need to be verified.