SatNav precision, accuracy or resolution...

[Externet]

Hi.
Do not know which term is the correct to denote how near a Satnav can pinpoint a location.

Checking specifications on the net, most vendors do not publish such figure. A few say up to 3 metres.
Are the newer units more precise, or that depends from the increased number of satellites launched trough the years ?
Were the older units uncapable of such precision or any GPS receiver can output enough decimals to behave as the new ones ?

Miguel

[MicroRem]

Hi Miguel,

I can partially answer your question, maybe someone with more technical knowledge can fill in the blanks. This info is anecdotal and may not be 100 percent accurate but I believe its pretty close.

When GPS was initially conceived, it was intended for military use only. Primarily US Navy, but the Air force and then the Coast Guard jumped in quickly too when its abilities became obvious. When civilian receivers became a possibility, the Navy began inputting an error into the system called "selective availability" which was an intentional error of up to a couple hundred meters to prevent adversaries from having the advantage of precision navigation. In order to receive the precision signal, a classified code was need to get the 2 meter accuracy the system was designed for.

The US Coast Guard then developed " differential GPS". In this system, a GPS receiver at a known site was coupled with computer and the local error was determined and rebroadcast to the local geographic area as a subcarrier on the navigtion beacons that were currently in use. This allowed GPS use in harbors and such with good accuracy, but the system was pretty bulky and need two receivers, one with a low frequency antenna (read big). It worked ok on ships with lots of room, but wasn't adequate for cars and such applications.

In the mid 90's an executive level decision by the Commander in Chief directed the Navy to disable selective availability and that is what made the hand held or dash mounted navigation units practical. The idea that someone would use precise navigation info to target the US seemed unlikely, I guess. The Navy can switch Selective Availability back on if our defense posture makes it appropriate. I assume this would require a presidential order but don't know that for certain. It is also unlikely we would know we need it until after the fact.

Also, as you noted above, there are a lot more satellites in the constellation now, in the early days we had to wait for a window in which we got enough birds for a good fix, now there's a bunch up there, and yes, precision is enhanced when you have more satellites in range.

Sorry for the long winded answer, I hope some of the info is helpful

Tom

[dyarker]

What Tom said, except the Air Force manages GPS, the Navy managed predecessor navagation systems.

Cheers,

[CeaSaR]

Having been in the land surveying business for 26 years, I've seen the rise of radio satellite positioning,
or GPS, come into widespread use over the past 10 years - not just in surveying, but other aspects of
construction as well. Most notable is the use of 3D software to control just how precise a Cat D9 can
blade a stormwater management basin while being semi-controlled from the sky.

As MicroRem (Tom) pointed out, the type of accuracy provided for this type of construction is on the order
of about 1 cm using a GPS base station on site to locally correct what is going on around it. The first step
is for a Surveyor to come out and establish a couple of high accuracy field points around the edge of the
worksite. Then the crew (whomever may be used by the site contractor) will come in and shoot those points
with a level 4 or 5 GPS receiver (could be occupied for 1/2 hour to several hours, depending on the required
precision) and that data is taken back for conversion to a local coordinate base. During this time, someone
with the contractor will build an extremely sturdy mount for the base station, usually next to the main job
trailer. The base will occupy this mount for the duration of the job, usually only being taken in when there
will be a protracted absence of workers from the site (long holidays). Whenever the base is "moved" from the
mount, it must go through a check to verify that it is in the exact same 3D location as before. Most always
it is.

To testify to the ability of the system, the firm I worked for had to verify 1 stake on the corner of a 150'x400'
building pad that was being filled (lots of large machinery, tracks, rubber tires, AND rollers) because they said
that it was out. We ran it back in using our total station using optical sites and stationary prisms. With all the
work that had been done around it, the stake WAS off about 1" east by 3/4" north. I can't remember how far
off the elevation was. After it was reset, the D9 operator placed the corner of his blade on it and said that we
were still 1/2" off for elevation, but within 1/8" north and east. We double checked all of our setups, angles,
distances and elevations and we were dead nuts. I'd say that 0.01'x0.01'x0.04' (surveyors/engineers measure
in 1/100 feet) on the corner of a slightly worn D9 blade is dad-gum good.

As to how well the systems work outside the construction arena, sub-centimeter accuracy is common. Greater
accuracy (<2mm) can be obtained with clear site at as many satellites as you can see and longer wait times
at the waypoints. There are systems that you can mount on your survey vehicle that can record 0.5' or better
accuracy while driving 55 MPH down the road. The potential for high accuracy is there.

Of course, what I am relating is of the professional grade and comes with a professional price. Last I checked,
entry into "the club" started around $35,000 for a base, backpack unit and post processing software. If you
wanted to up the ante, $70,000 would get you 2 more backpacks and a subscription to a local GPS net (verified
base station run by a company) for a year.

The consumer grade handheld (Magellan type) units used by geocachers, hikers, bikers, etc. are in the level 3
(consumer) band that ranges from ~3 meters up to 20 meters, depending on the precision afforded by the
receiver and the situation you find yourself in (how many satellites you can obtain). This is generally good enough
to keep yourself out of trouble in the wild.

Hope that helps.

CeaSaR

[MicroRem]

Ceasar and Dyarker, thanks for the additional info. I appreciate the info and will file it away in my brain to recall it incorrectly in a few more years. The precision Nav stuff is really world changing technology. In my early days in the Coast Guard would have never even dreamed about this kind of accuracy. GPS was just getting firmly established as the standard when i retired. I always worry though, that crews will get so used to GPS and plotters with overlaid radar and they will be clueless when the display goes out and they have to go back to a paper chart and dead reckoning, lousy fixes, etc, that defined most of my career!

Have a great Thanksgiving week!


best

Tom

[jimmy101]

In addition to the "intrinsic" accuracy of a GPS receiver you can often increase the practical accuracy beyond the intrinsic accuracy. And, you can increase the precision even farther than the accuracy.

Back when civilian systems were limited to ~10m resolution it was possible to get down to a resolution of a couple meters or better if you had a large enough area to work with and could take multiple readings over a grid that was a couple times larger than the GPS's resolution. From the 10m data it is possible to interpolate down to perhaps a meter or two accuracy and probably better than that in precision.

The same technique is often used for aircraft and rockets and such. The movement of the craft makes it possible to interpolate position to a greater level of precision than that of the receiver.

Even in a car, if the car is moving, or is moving up to a particular point, the stream of GPS data can be used to calculate the position to a finer resolution than what the receiver can do if it is stationary. (I don't think car receivers do that though.)

[haklesup]

Simply put, Accuracy is how close the GPS reciever can fix your location WRT to your actual and absolute location. Resolution is whether the reciever can correctly report that you have moved 10m (or any distance) when you actually do move a measured 10m. (Accuracy=Absolute, Resolution=Differential)

Generally, resolution is better than accuracy but both are effected by the quality of the signals you recieve (how many satelites and if any are obstructed view and multipath noise). With GPS, X and Y is always more precise than Z. Accuracy and resolution can suddenly change if a satelite moves in or out of view. With high resolution but low accuracy you could say that you moved a certain distance but that location might be offset from a map actual position. Often with older handheld units you would see your breadcrumb trail appear next to the road you were travelling on. Newer units correct for this by forcing the icon to be displayed on the road. This is annoying if you try to use one off road. Before SA (Selective Availability) was turned off, you would see the breadcrumb trail actually move an offset distance spontaniously or your position jump around even though you were stationary.

Alternatively from the POV of the reciever, Resolution is the smallest increment that the instrument can resolve even though the available data (the signal from space) is less accurate. A reciever may be able to resolve <<1m under ideal conditions but only 3m under typical conditions.

Industrial units improve resolution and accuracy by using base stations with well known coordinates to act as a local sattelite. Military units achieve better specs by decoding portions of the sattelite signal that are scrambled to all but military grade recievers. The government reserves the right to turn SA back on for security or defence reasons without notice. For example if they ever thought terrorists had smart bombs, it would be turned back on indefinately.

[kheston]

Ceasar,

Does the Trimble and Thales survey stuff that comes up on eBay all the time even hold a candle to the pricey equipment you're talking about? Any idea what level of accuracy can be had for the $500-1500 range (used) that this stuff seems to sell for?

[CeaSaR]

Those are 2 of the big trade names out there, but for that price you are probably looking at at least 2 generations
behind the current level. IIRC, there was a level that the goverment was going to turn off in favor of a new one and
these units may be on that level, effectively making them useless. Of course, without the software to run them
and interpret the data, they are useless also.

Hmm, the prices seem way low for this type of equipment. I'd ask for serial numbers and confirm with the manufacturer
about the status of the piece (legal/stolen, still on an operational level, etc.) before even considering it. The original
owners would normally register them for insurance purposes and subsequent owners would most likely do the same.

As for their accuracy, they are right up there with the new stuff, it just takes lot longer to obtain it. The new stuff is
supposed to be much quicker and more tolerant of conditions.

CeaSaR