Self-driving cars have been dominating the news cycle lately, but NASA is hoping to take automation even further. And because it’s NASA, they’re shooting for the stars: Instead of self-driving terrestrial vehicles, NASA is hoping to develop exploratory satellites that can learn how to navigate by themselves.
Although satellite technology has continued to develop steadily, one issue hasn’t been satisfactorily addressed: satellites still require wireless transmissions from Earth to be controlled and repaired. As a satellite heads deeper and deeper into space, this can pose a serious challenge. Currently, scientists on Earth do their best to predict, for example, where space debris might be, or when a satellite has the best opportunity to collect data, and they then send transmissions to the satellite with instructions based on those predictions. It’s a laborious process.
Now, NASA is hoping to develop satellites that can make some decisions “on the fly” by utilizing AI, fuzzy logic, and a version of the Ethereum blockchain network. To move research forward, Nasa has awarded a grant to the University of Akron, where associate professor Jin Wei will lead a team devising what the NASA press release refers to as a “resilient networking and computing paradigm.”
According to the University of Akron’s own press release, Dr. Kocsis’ current research “aims to develop digital thought that is data driven and emerges from multiple thought centers, rather than a single central processor,” which would theoretically allow a satellite to tackle multiple challenges simultaneously. In the future, this could allow satellites to autonomously navigate through orbiting space junk and make its own decisions about data gathering and processing.
The funding, one of a handful of Early Career Faculty for Space Technology Research Grants, totals $333,000 over three years. Along with a press release, NASA also released a Venn Diagram to illustrate the overlapping technologies that they and Dr. Kocsis hope to synthesize, which can be viewed here.
Image via AdobeStock