> "Well, currently GPS can estimate your location to an area of about 20 feet."
With DGPS we get about +-10cm for our non-military robotics project. So this isn't really true. We use a paid subscription but the free WAAS signal brings the accuracy to about +- 1m which is already pretty friggin good.
I wonder what justified this (probably expensive) upgrade?
The existing satellites have been in service too long. It is a "replace it or lose it" question.
It took me a while to find out why they are more accurate, I eventually found this quote: Greater accuracy is a product of advanced atomic clocks in the new satellites, Torok said. I think it's safe to safe that tiny atomic clock technology advanced in the decade between the existing satellites and this round.
They added a new military signal, presumably jam resistant. (It has a "spotlight" antenna that can provide a 100 times more power to a fighting theater. I'll see your alibab.com gps jammer and raise you 20dB…)
They also added another civilian signal (L5 - safety of life) which presumably has a purpose, but darned if I can find it. Probably related to different frequency, more power, different coding, something about the ionosphere… my guess is that it won't crap out at the same time the other one does from interference.
The L5 moniker "safety of life" is an accurate name. It will assist GPS aircraft navigation, especially during the safety-critical landing approaches. Aircraft rely heavily on both GPS and WAAS satellites for corrections. Actually, L5 has been broadcasting on the two WAAS satellites for several years, but most end-users would not have used it. AFAIK, L5 decreases the error from multipathing and the ionsphere, among other improvements.
You're using ground-based assistance to get that extra accuracy. This upgrade gives the satellite signal better accuracy without the need for ground based assistance stations (and expensive receivers).
I know, that's the point of DGPS. What I meant was, we already have that infrastructure and accuracy, so why are they putting new satellites? I guess the ones up there are getting quite old now.
Are your robots moving around? I'm just curious because the positioning research I've been exposed to couldn't get better than ~50cm accuracy for objects in motion (for example, a person walking around at a normal pace).
He's subscribing to Omnistar HP service. Basically, it's a float RTK solution relative to a synthetic base station. They boast 10 cm error 95% of the time. It's expensive (IMHO) and breaks whenever sun flares up.
True RTK + IMU positioning systems will get you 1-2 cm 95% of the time (see Applanix, Novatel).
Interesting! I don't use this stuff in anger, so my questions are merely idle curiosity.
Omnistar refers to the HP accuracy as "real-time"[1] and:
> [it] will usually have a 2-sigma (95%) horizontal error of about 6 centimeters and a 99% horizontal error of less than 10 centimeters
Do you happen to know what kind of movement this is based on? For example, if it was attached to an RC car (assuming that the equipment is this compact, it may not be?) would you still be able to get this level of precision? Or is it designed for more consistent movement patterns?
My guess is that it doesn't assume anything about your motion model. The real question you should ask is what's the frequency of the positioning solution. Your generic GPS unit computes solutions at 1Hz, more sophisticated ones do it at 5Hz, surveyor quality units can do it at 10-20Hz depending on firmware, but I think they're pushing the limit and probably assume something about the motion model. Applanix and Novatel solutions that use inertial sensors definitely make pretty strong assumptions of the motion model (see Kalman filter).
OmniSTAR in particular and DGPS in general will operate under assumption that errors due to ionospheric and tropospheric effects change slowly in time and space. If this assumption is correct, then dynamics of the vehicle won't affect the corrections themselves.
My point with all this is that 99% of the corrected GPS solution should theoretically be within x cm of some point on the true path of the vehicle no matter how it moves in a plane. The question is, if you're interested in real-time solution, how far will the vehicle be from that point by the time you get this fix, and if you're interested in capturing vehicle's travel, how well will the curve fit through these points match the true vehicle path.
There are other ways to get even greater accuracy than an applanix. But they require pre-indexing the surroundings and using a lidar to position yourself within the "virtual" world.
Haven't worked with Riegl, but I can tell you that Velodyne isn't even close. Even their official datasheets state range error of 5 cm; reality is much worse.
The factory calibration on the 64 diodes of the velodyne is bad. It needs to be fixed. There is a paper out of Stanford that tells you how to fix it. I can't think of the name off the top of my head but if you email me I can dig it up.
Also the error grows with distance. So if your bot is navigating off of objects 200 meters away, yeah you'll be more off. But you can use points closer to it and you'll get better results. Look around and you'll see that the velodyne has been made to work. Nobody loves it though.
The riegl has very low beam divergence, is beautiful, survey accurate, not waterproof and $300k+.
Don't forget the workhorse SICK scanners. They are cheap and have their pros and cons.
It would be nice if accuracy could be traded off for time, so that instead of getting a more accurate lock, we're getting a the old accuracy but quicker.
For me the biggest limitation of GPS is that it can take > 1 min to get a lock on a regular basis. I know this is hardware and software dependent ... but it's still common for it to be at least 20 seconds even on the best hardware, I think. Apart from making apps that rely on GPS awkward ("please wait an indeterminate amount of time while we find your location") it also means much greater power use because the GPS has to be active for longer (GPS seems to be something that noticeably uses a lot of power in my Nexus One).
While too advanced for the phone at this stage, there is something called "direct inertial aiding" that solves this problem. Basically it uses the gyro to estimate the position during a gps outage and then it can tell the GPS where to look in the sky for satellites. As it is, the gps has to scan the entire sky and it assumes nothing about its position. Even if you just passed under a bridge and only had a monetary outage. Direct Inertial Aiding makes pick-back-up times very small.
A lot of the delay in getting an initial GPS fix is downloading the almanac for the satellites at the very slow GPS bit rate. The iPhone 4 probably periodically retrieves the almanac over GSM instead, allowing much faster location fixes. (Several models of Android phones do this.)
My GPSmap62st sometimes figures out where it is on my desk indoors 3 m away from the nearest window. Accuracy is horrible (it often thinks it is outside), but it averages to the right place eventually.
And there's another floor above mine, as well, complicating things. I'm quite impressed it is able to receive anything, but it just goes to show that with the right antenna and DSP, amazing things are possible.
Yes, but the receiver clock plays a large part in accuracy (as does software). With the satellites travelling at 14,000km/h, even being accurate to 10^-6 seconds means you can be off by 280-300m.
This new series of GPS satellites will have more accurate atomic clocks, and some other changes (such as being upgradable remotely). So I doubt older devices would be a lot more accurate since they still have clocks that are orders of magnitude less accurate than the clocks on the satellites.
With these fine measurements, even the smallest delay or inaccurate measurement in software or hardware can make a big difference.
There is a better technical description of the new series here:
With DGPS we get about +-10cm for our non-military robotics project. So this isn't really true. We use a paid subscription but the free WAAS signal brings the accuracy to about +- 1m which is already pretty friggin good.
I wonder what justified this (probably expensive) upgrade?