The Air Force Wants to Use Earth's Magnetic Fields as a GPS Alternative

If so, those sensors would be capable of locating the plane based on where it was in relation to our planet's own magnetic fields. His experiment showed that the method was a viable option as a GPS alternative.

But just because the method might be viable, doesn't mean it is, however, easy or straightforward it seems. The electrical operations of the plane itself interfere with a sensor’s ability to detect the strength of the field.

This is where the Air Force decided to use artificial intelligence. AI is famous for canceling noise from sensor readings. This results in a better more accurate signal.

Researchers in the Air Force’s-MIT Artificial Intelligence Accelerator community, joined forces with scientists at MIT to work on noise cancellation. They published their own paper in July.

They found that magnetic field readings can be accurate to ten meters. Compared to GPS readings, that are accurate down to three meters, this might seem like a real disadvantage. But magnetometer readings have one key advantage: they are much harder to jam.

“Because of the size of the earth and the magnetic field… it takes a whole lot to jam a signal coming from the earth, and by a whole lot I mean on the scale of a nuclear blast,”  Maj. David "Stitch" Jacobs with the accelerator told Defense One. 

“Apart from that, it would take a giant scale of a machine to block what’s coming from the earth’s crust. But then you could also cancel it out with machine learning”

The Air Force in collaboration with MIT is now reaching out to the AI community to seek out better tools for cleaning up magnetic field readings. They have released a new joint accelerator program that ends August 28.