The model doesn’t just help you get from the office to Taco Bell: Ships and planes–civilian and military–also rely on it. “One of NGA’s biggest customers is the military,” says Paniccia. “If you’re sailing across the ocean in an aircraft carrier, it’s very important you know where you are and you’re not going into enemy territory.”

That requires information from the magnetic model, and hence uses data from the Swarm mission. Swarm uses satellites of the old-school sort: giant, expensive, full of lots of sensors measuring lots of different things, including but not limited to magnetism. “It was not launched for the purpose of collecting magnetic data,” says Paniccia. It’s currently funded through 2023 (and may get a life extension), but it won’t last forever, and it doesn’t belong to the US. So now is the time to start thinking about what a newer, better, potentially US-based mousetrap looks like.

“We’re looking for what’s the next best way to get the data,” says Paniccia. Is that small, specialized satellites? Tiny sensors on the ground? “That’s where this MagQuest idea comes from,” he continues. “Let’s open it up.” The agency staff want to see who–at a university, in a lab, or at a private company–has got a big idea.

When the agency opened the competition, officials weren’t sure anyone would have any ideas. “Our biggest worry when we started this was we’d get zero submissions. Or two,” says Paniccia. Instead, during the first phase, they got 40 competitors, all of whom sent in descriptions of the systems they’d like to construct.

The 10 winners of that phase each got $20,000–with no stipulations or earmarks, just cold cash. In the second phase, competitors had to drill down on the details of their instruments. They had to produce detailed designs and plans for how they would collect data, including what their sensor would be like, what platform it would be on, and how they would analyze the data. How would the system perform? What were its risks? And how might the team manage a future program? Based on those schemes, five winners split $1 million total.

Now, in the just-announced Phase III, innovators will bring their polar visions even closer to reality, vying for a $900,000 prize. The NGA isn’t under any obligation to buy the winning technology, or any magnetism-measuring technology, after the competition. But it may. “We have planted the seed that at some point in the future NGA is most likely going to put in a formal procurement for something,” says Paniccia. A winner of MagQuest would likely have a leg up in the quest for that hypothetical contract.

One of the teams, based at the University of Colorado Boulder, is planning to construct a small satellite: 10 centimeters wide and high, and 74 centimeters long, like a high-tech hot dog. That length isn’t for looks. The device that will measure the magnetic field–a magnetometer–will go on one end. The rest of the setup goes on the other end. That’s because the gear–like the metal on the long-gone boat–could mess up the magnetic measurements. Keeping the parts away from each other makes the data cleaner.

And keeping the whole apparatus small and uncomplicated–spacing out the instruments but not using a robotic arm to do so, for instance–is meant to appeal to NGA’s goal. Stuff that goes to space doesn’t live forever–radiation degrades it over time, for instance. So sometimes your best bet is to build clones that you can just keep launching. “If we’re going to have a solution that’s going to last for decades, we’re going to have to replace it,” says Boulder’s Bob Marshall, a professor at the Colorado Center for Astrodynamics Research and a leader of this MagQuest team. Small, simple satellites like this are inexpensive(ish). While it’s not trivial to send up reinforcements, it’s not nearly as costly as launching another Swarm.