Reading the article, it seems the real use case (unmentioned) is on ballistic missile nuclear subs.<p>You want them to be well below the surface of the water, where the GPS signal does not penetrate so well. Thus having a way for them to know precisely where they are without relying on GPS would be very helpful.
Neat. Geopositioning in general is a "hard problem" especially with additional variables/complexities that are introduced for specific use cases (military or not). There's tons of room for cutting edge research here still, and it's exciting to see companies invest in that.