There’s been talk for years of a more accurate Global Positioning System. The current GPS system tells you roughly where you are, but it’s only accurate to within a few feet. That vagueness means that although it’s fine for mapping, it isn’t good enough for narrowly targeted proximity or geo-fencing that can be used in e-commerce.
Existing GPS has been used in toll-road billing, and has been fine-tuned for surveying with large, expensive antennas, but it’s currently not much good for tracking customers as they choose a concert seat, for example.
The European Space Agency is building a new, highly-accurate system called Galileo that they say will be fully functional by about 2020.
Galileo, if you scan its fine print, includes authentication. That’s an accurately timed, trusted location factor that will allow advertisers to not only pitch you down to the micro-location, but also let you perform financial transactions.
You’ll be able to choose your seat at an open-air sporting event and pay for it too. Turnstiles become redundant, among other things. You should even be able to place your hand in a vending machine and get billed.
But, as I say, this is a ways away and Galileo still has to get more of its birds into the sky for it to work properly—a 2014 launch placed two satellites in the wrong orbit, which hasn’t helped progress.
In the meantime, a University of Texas at Austin project to improve GPS accuracy using existing tech is definitely interesting.
The researchers say that they’ve gotten GPS location errors “from the size of a large car to the size of a nickel – a more than 100 times increase in accuracy.”
And they’ve done it with software in existing smartphone-quality chipsets, antennas, and the existing GPS constellation.
The process involves extracting accurate data, called carrier phase measurements, carried by GPS anyway, and used by surveyors and special interests, like mining. Previously, collecting those accurate measurements required special antennas.
UT Austin’s software-defined GPS receiver ideas don’t have an authentication factor like Galileo, but do have accuracy and use GPS’s global reference frame. It could be in a small form factor, ultimately.
That improvement in accuracy, with cost savings and size, could make the system suitable for automobile collision prevention, outdoor virtual reality headset games, the Internet of Things, and drone deliveries.
The Cockrell School of Engineering scientists at UT Austin say poor multipath suppression in existing smartphone-grade gear is causing ambiguity in processing, and that degrades accuracy.
Multipath is where signals travel along many paths—not always for the best results. It’s what caused ghosting on old analog television broadcasts, if you remember those. Reflection can cause multipath.
The scientists say that it isn’t the antennas, which is what was previously thought to be the impediment. In fact, they say that chips in smartphones can be better quality than those found in costly surveying GPS receivers.
You don’t need expensive, bulky gear to get accuracy, they reckon. You can do it all with software and a form of random antenna motion where the smartphone is gently moved in wavelength-sized increments to reduce multipath.
The researchers have written about their tests and findings in a GPS World article, if you want to read more.
Next up is to figure out a way to further reduce the error-creating multipath—possibly by estimating trajectories for the accuracy-enhancing antenna motion.
A snap-on smartphone accessory is currently being developed by the group, which created a Samsung-deal financed startup called Radiosense.