DARPA considers GPS unreliable, as it isn’t always available and signals can be easily jammed.
What takes this into a step change from evolution, is that’s what it is, is the use of ordinary every day objects. Whether it is revolutionary or evolutionary may be something they don’t share for now if it is about getting an edge on military opponents.
A key issue is the cost of technology. When I got my first in-car navigation system, the NASA designed gyros and accelerometers which gave me high accuracy put the retail price tag at around US$4,500. I didn’t lay for it, part of my job was the launch of these systems in New Zealand.
Our mapping car however had a system which was also large and bulky and in the region of $200,000 and the reason was that it had to be accurate to within 15cm even when no GPS was available, because being down at the bottom of the planet, the spread of GPS signals was not designed to cover all of our country down at the bottom of the planet.
Today the use of assets or Points of Interest can be valuable, but of course there are questions about how permanent those assets are. Systems like iBeacons make a lot of sense. After all it worked for Hansel and Gretel. Passive RF devices can be placed pretty much anywhere, they can be dropped on or into dirt, placed on buildings and can provide breadcrumbs anywhere. As long as you have the technology to know where they were placed, they can work for a long time because some of those technologies only consume power of significance when they are activated.
When they talk about TV, I wonder if that is designed to put us off the scent or whether they are talking about Internet connected or smart TV’s because they, being effectively WiFi access points therefore are transmitting a unique signature.
Google got itself into hot water by tracking WiFi signals from people’s home WiFi transmitters which also have a unique signature and aren’t likely to move very often, therefore if DARPA uses technology to map the location of those access points (they don’t need to spy on what is being communicated across them, all they need is a unique identity of a fixed router, whether it is in an office, a public device on a street or in a home. Triangulate those unique mapped signatures in an urban area and you will have pretty good accuracy. Telcos have been doing this and improving on it for years. It’s much cheaper for them than GPS, which requires either legislation or informed permission from consumers, which is typically only given for the purposes of the apps they are using. Then there is GPS assisted tracking which is probably what DARPA are using. It doesn’t replace GPS altogether because there are inherent safety elements to GPS and typically other technologies, such as those we used in our mapping cars were still based on starting each day with a highly accurate GPS fix. From there we could function without it for long periods of time. More satellites are going up and jamming them or using EMP bombs on a satellite is likely to be more difficult.
The consumer will ultimately benefit from these technologies if we are serious about driverless cars. Cars that navigate themselves without a human sanity check solely based on GPS will fail, even in urban canyons where the signals bounce of he glass and metal foundations on urban buildings. Have you ever sat at a red light in a city and had your nav act as though you were still moving and instinctively pressed a bit harder on the brake pedal?