Why self-driving cars need satellite guidance to take off
Search Driven for for sale
Autonomous cars need close guidance and the SBAS technology aims to do that.
Self-driving vehicles are the future, everyone agrees, and clever sensor and camera technologies are being developed to stop them from crashing into each other and ensuring as few people as possible are hurt when accidents are unavoidable.
Nevertheless, autonomous cars need to know not only their destinations but also how to safely get there: they need guidance, usually from high above, via a global navigation satellite system (GNSS). This would be the United States-developed global positioning system, or GPS, Europe's Galileo, Russia's Glonass and China's BeiDou.
There's a problem with GNSS delivered from the skies though, in that the service isn't terribly accurate.
Finding the correct spot within 5m to 10m might be okay for people who can work out themselves where to go after being given the rough location. For moving machines that can't make educated guesses like we can, such low accuracy would be a disaster waiting to happen and preclude self-driving cars outside cities.
More accurate positioning systems are in the works, however.
The Australians decided to throw some money on building a satellite based augmentations service (SBAS) in January this year.
New Zealand has now joined in, adding A$2 million ($2.2m) to the A$12m for the two-year SBAS trial, which has now kicked off with test signals already being transmitted.
Depending on the technology used, SBAS could provide positioning accuracy down to 5cm, although the trial will also comprise a single-frequency augmentation system that offers 1m precision.
The idea behind SBAS is to use existing GPS signals and improve their accuracy with ground-based infrastructure.
Besides the single-frequency service, the New Zealand and Australian trial will test next-generation SBAS. This uses dual-frequency signals and multiple satellite constellations for better positioning performance.
The precise point positioning (PPP) system will also be trialled, and that's where things get really accurate, to within centimetres. PPP may not be suitable for real-time applications though because the service takes a long time to initialise.
Land Information New Zealand, Geoscience Australia, and the Co-operative Research Centre for Spatial Information will handle the SBAS trial, with US aerospace giant Lockheed Martin and Inmarsat satellites, and GMV for the ground-based equipment.
As the SBAS signal is already transmitting, you may find that your positioning system on your device is already improved - but the test bed is not yet certified for safety-of-life use, of course.
Although one goal of the test is to determine if Australia and New Zealand will pursue the development of an operational SBAS, it's hard to see how this wouldn't happen: it's not just self-driving vehicles that need greater positioning accuracy, but other applications as well such as livestock tracking, finding ships and sailors in distress at sea, and much-improved logistics systems.
We're a bit late to the SBAS game, in fact. The United States already has the wide area augmentation system (Waas), and across the Atlantic, there's the European geostationary navigation overlay service (Egnos).
Russia, India, Japan also have SBAS infrastructure to correct less precise GNSS systems.
It's good to see that we're catching up with the rest of the world on what is crucial technology for immediate-future applications - we literally would not find our way without more accurate location systems, and we'd fall behind the rest of the world.
-Juha Saarinen, Tech blogger for nzherarld.co.nz