Smart space habitat looks to put AI to work in deep space
Researchers at Carnegie Mellon University are working on a smart space habitat for deep space missions. Part of NASA's US$15 million multi-university Space Technology Research Institute's Habitats Optimized for Missions of Exploration (HOME) project, it's purpose is to help design advanced autonomous systems that use artificial intelligence to monitor the habitat and anticipate and automatically correct faults.
So far, space agencies have been very lucky when it comes to living outside the Earth's atmosphere. The farthest expeditions were only a handful of Moon missions that lasted less than a fortnight, and since then astronauts have bunked down in low-Earth orbit space laboratories where home is only a few hours ride away – not even far enough to worry about a case of appendicitis.
But as astronauts venture farther out for longer periods of time, the reliability of their habitats becomes of primary importance. Such systems, no matter how much redundancy is built-in, can't be replaced by picking up the phone, so they must be constantly monitored and fully understood if catastrophic breakdowns are to be avoided.
According to Carnegie Mellon, one answer to this is artificial intelligence. Modern computer algorithms designed to seek out and recognize patterns can take on important tasks like watching electricity usage and use it to determine the system's status. It's a method that's already seeing applications on Earth, but these often require a large amount of data to work – something that a space habitat may not have.
"How do you conduct automated fault detection and diagnosis without a lot of system data? This is where AI comes in," says Associate Professor Mario Bergés, who heads up the research team. "We have machines that learn by themselves if you give them enough data, but we don't have a lot of machines that can reason by using existing engineering knowledge, which can reduce the amount of data they need."
As part of their study, the team will take a series of electrical measurements of an experimental system and find out how to use AI to produce meaningful results that can then be either fed to robots to make corrections or as advice to the human crew. The hope is that this will prove useful in the design of NASA's Gateway deep-space outpost.
"Since the beginning, civil engineers have been the stewards of the infrastructure that supports modern life," says Bergés. "If humanity is moving into space, it makes sense for civil engineers to be part of that."
Source: Carnegie Mellon