Smart Phone GPS accuracy

> I'm working along a UDOT road. Taking pictures of the ROW markers as I RTK the position. I've got the location turned on in the phone so it records the location. So which is more accurate?
>
Resolution and Accuracy are not the same. Depending on the GPS chip in the phone, you might have resolution greater than that. Doesn't mean it's where it says it is.

You could record the wrong position to as many decimals as you'd like; you still might not be in the right place.

I say RTK with OPUS is more accurate than most any phone.

Then again, if the question was tongue in cheek, and I've taken it far too seriously, please forgive.+o(

Reported Decimal Places Do Not Guarantee Accuracy/Precision

Your phone is using as few as 4 single frequency GPS signals from a cheap antenna and broadcast satellite orbit positions to calculate coordinates without any thought to atmospheric correction. Phones may be limited to 4, 6 or 8 channels.

Meanwhile your RTK is using double frequency or more (atmospheric correction) from a survey grade antenna (electronic filters against multipath, etc.) and comparing other corrections from a precisely known fixed point to give a precise and accurate position.

Seriously, if you have to ask this question you may not be qualified to do precise positioning with any type equipment.

5 decimal degree places and +/- a lot for a phone, 8 decimal degree places +/- 0.02' for RTK.

Paul in PA

If we use 3959 miles as the mean radius of the earth, 1 second of arc would be about 101 feet. So, the 14th decimal place would be 101 * 10^(-14) = 1.01 * 10^(-12) feet. Along a parallel at your latitude, the number would be about 7.7 * 10^(-13) feet.

The phone's decimal coordinates convert to N 39d 39m 10.48370s, W 111d 27m 1.99402s. Using the lengths from the mean earth, the phone coordinates are different from the OPUS coordinates by about 19.53 feet in the NS direction and about 4.56 feet in the EW direction. By Pythagoras, that's about 20 feet away.

That's a great point that you've made. Implied accuracy is very different from accuracy. Thanks for sharing the data.

Well, considering the obvious problems with RTK and its pervasive, irresponsible use by supposed "professionals" that loosely call themselves "surveyors", the position returned by the phone should be much more precise and accurate. However, there must be multiple observations per hour over at least a 48 hour period. Once these carefully obtained observations via "smart" phone are rigorously manipulated in Star*Net, they are good as gold.

Ignore the RTK and trust the Samsung/Star*Net results.

Oops. Correction. You're talking about the 14th decimal place in degrees, not in seconds. For the mean earth, that number is about 3.65 * 10^(-9) feet.

Either way, it's a bit less than my eyes can see.

The phone's coordinates appear to be good enough that I can sort out which picture goes with each row monument. Why the phone lists the decimal lat and long to 14 decimal places seems sort of crazy to me. If the accuracy is in the 20 feet range they should cut if off at maybe no more than 1/10 of that or 2 feet which is still overkill (in my opinion).

RTP (real time phone) enhanced version by Star*Net, with county GIS coordinates, using decimal lat and long to 14 decimal places, who needs a surveyor anymore?

RTPS*Ne^14-GIS enhanced, only way to go! Anybody with a smart phone can do it!

How much more accuracy might one expect for a smart phone than a regular car GPS? My car GPS typically gets me within let's say 10 feet of my destination. My folks have a GPS that gets them within a couple hundred feet.

If I type my address into Google Maps, I wind up about 3-4 houses up the street. In other places of course, the address is more accurate.

I believe Google stores an address for each end of a block or series of blocks and does a linear interpolation between those points for the location of all the in-between addresses. So if your street isn't linearly spaced or they entered the endpoint a little off, then everybody in the neighborhood can be shown at the wrong location.

It's definitely overkill and misleading to non-numbers people. We could only speculate as to why so many are displayed.

One answer may be that it's just easier to print the raw number than it is to round it. Sometimes, negatives cause rounding routines to produce wrong answers.

I often show more significant digits than necessary or justified so that students can duplicate my calculations unambiguously, but I doubt that that's the case here.

The important question is "How accurately were the ROW monuments placed?"
One of the most important aspects of land surveying.

I have argued with GDOT about ROW monuments not being set by licensed surveyors. GDOT allows the roadway contractor to set the monuments and then check them by (as one inspector told me) "we count the number of monuments, and if that number agrees with the number that is called for in the plans, we sign off on them".

But in response: constrained measurements (RTK) are inherently more accurate than "free range" (cell phone) measurements.
I would liken the two to measuring with a chain and having a rear chainman (RTK) and dropping the chain on the ground near the pin, then pulling ahead (cell phone.

Same problem in TN. I've always joked that contractors just throw them out of the back of their truck.

Anybody remember the trimble submeter gps gis ad in pob a few years back that showed the coordinate decimals out to a ridiculas amount.

Smart Phone (w/peripheral GPS) accuracy

best smart phone gps story i think i can tell (and probably will ever be able to tell): went out to check the as-built of a detention pond on a car wash site about a year ago. pull up on a saturday morning, it's fairly empty (whole place was under construction), this jeff spicoli-looking guy comes out to greet me. he's the super, so i engage in the standard q&a with him, which at some point leads to "so where's your benchmark?" he replies that there isn't one, that it had been knocked out weeks prior. i'm looking around, thinking there's still a good bit of stuff needs to be laid out, and this is a car wash and drainage isn't exactly a back burner issue. "so how you holding elevations, then?," i asked. "oh, i got this iphone app, i keep checking a few hubs here and there with it and everything seems to be just fine."

i quit asking questions at that point. ran a couple of rapid static shots in, shot finished finished floor, shot the pond, ran a level down and back to something i knew was nearby just for grins. gave the engineer the survey and said "good luck."

At the eighth decimal point in degrees - 0.00000001 degrees = 0.00000000017 radians

so at 1 click at the eighth decimal place the distance in latitude would be:

arc length = angle (in radians) x radius

0.00000000017 x 3959 mi x 5280 ft/mi OR 0.00355 feet (0.003648 out of calculator without rounding)

14 decimal places, I just don't see the need for that!

Certainly, all of those digits are not usable, nor are they useful. On NGS Data Sheets, the fifth decimal place in seconds for lat/lon represents 0.00103 feet, using the 3959-mile average earth radius. Eight decimal places in degrees would state accuracy slightly less than NGS while nine decimal places would state accuracy slightly greater than NGS.

Since the instrument's numbers are about 20 feet different from an NGS-like measurement, it doesn't really matter. But the arithmetic is interesting and the question of how many digits to display should be considered with more care than this developer used.

It's also interesting to look at the data for Lambert states in James Stem's NGS 5 publication. The number of significant digits shown for the latitudes of the central parallels and other constants rivals the cell phone numbers.

ADOT used to perform in this fashion too, as well as setting monuments on "approximate section line", approximate 1/16 line," etc. It took a complaint from an RLS to the SBTR and an aggressive AG rep. to get their attention. They perform much better now.

Smart Phone (w/peripheral GPS) accuracy

H2O runs uphill in Texas- everything is better in Texas- per my second ex-wife from Dallas. (TIC) also "digital dummies".

I will not say I am glad that other DOT's have this attitude, but I am relieved that my home state is not alone.