Land surveying 101 - How accurate is RTK GPS
Вставка
- Опубліковано 5 лип 2024
- Today's video will discuss some important details to keep in mind when deciding if RTK GPS is accurate enough for the work you are trying to do.
00:00 - Intro
00:10 - Precision vs Accuracy
02:38 - What do the manufacturers have to say?
04:36 - Do RTK measurement errors follow a normal distribution?
07:25 - What causes errors in GNSS measurements?
11:04 - ISO Testing & manufacturers specs
12:49 - Ideal Conditions Testing Results
15:53 - Less than ideal conditions
18:37 - Distant base station
19:30 - Final thoughts
Datasheet download link - 1drv.ms/x/s!AkXoacH8d5pAhJoQK... - Наука та технологія
This channel is pure gold! Thanks for your effort!
Thanks for these videos. They are very useful and well explained. These are many of the facts nobody takes into account when using GNSS. I really appreciate your work and your videos. Thanks again
i dont know what else to add to the actual comments, great video and great explanations.
Thank you!
@@The3rdDimensionSurveying I have noticed now that the link isn’t working
If I could give you 100 thumbs-up, I would. Excellent video. Hands-down the best RTK primer for surveyors I've seen on you-tube. Bar none. There are too many right out of college 2nd year practitioners giving half-good advice here. I'm happy to say that this video is nothing of the sort. Let me detail why I can say that and why my opinion might matter a little bit.
I've been active as an RTK surveyor since the mid 90's, built and manage a GNSS network, trained surveyors for 2 decades, spoken at our state conference on RTK precision, help with standards development for a large state DOT, evaluate new systems before we purchase them and many other aspects of this work in my 33 year career. I've been doing these test you show for 20 years. I got fed up in the early 2000's when I couldn't get a straight answer, just like you described, so I started testing myself. I am constantly amazed at how well RTK works, as well as how it can hurt you when you trust it too much. My findings match yours very closely and really haven't changed much in the last 20 years. Around 2008 or 2009, I saw a slight bump in vertical precision when we started using RTN. I estimated that it approximately halved the PPM error, and sure enough, they ended up putting that in the specifications years later when the Trimble R10 was released. That was very satisfying.
I summarize it like this (Basically what you said!!): we can generally take the specs shown as usually good to 2-sigma as long as we have low DOP and good atmospheric conditions. However, probably 1/20 - 1/50 of the measurements will be way worse than all the others, and there is no way of knowing which those are unless we shoot them again later. Also, this assumes that we have well maintained and calibrated equipment and good procedure and adherence to those procedures. Once you get into a less-than-ideal situation, all bets are off. I will say that tightly coupled inertial systems have the potential to change this dynamic under certain conditions (over a short period of time), but I'm now in a managerial role and don't supervise field crews, so I don't think I'll ever be testing that .
I've run so many spreadsheets with custom stats and graphs. I've lost more than I can now find. One thing that opened a lot of people's minds was when I did a test of 'dumping' the receiver and re-measuring, which is something that I always have argued against, despite it being one of the favorite things sales-people showed us to do in the 90's when these were being rolled out. I was able to show the positional bias over a short period of time, proving that you must have a time separated measurement to get a normal distribution. I believe you shared a similar finding.
I've also been combining RTK, static and terrestrial measurements for a long time and am constantly frustrated that it isn't done more, even in my department. I can't stand that people still shoot in RTK 'pairs' and then "lock 'em down" before starting a traverse. Why hold 1 elevation over another? Just hold them both and run a least squares adjustment (I use TBC and TGO before that).
The way I describe the precision/accuracy is this. RTK will bring you precision, your constraint to your control and its quality brings in the accuracy. For years we would just occupy passive grid coordinates and do check shots. When we started using our RTN in the mid 2000's, we started doing site calibrations/localizations. The latter gives you a pretty good idea of your survey accuracy. Then, if you shoot 1 point on the site about 15x and run the stats, you can get an idea of the precision at that site. But that is only for the same person, using the same equipment, in the same conditions.
I really can't talk about when we are going to start to reign in the non-professional activities that some are using these tools for, but a movement is afoot in my state. It isn't the tool, but the use that is the issue. The first thing that needs to get cleaned up is we surveyors have to be better at using this tool Your video here a good start, so thanks for taking the time to share this. I never had much interest in editing videos, so I do appreciate the amount of work that went into this. You can DM me if you want any additional information from me.
By far the best comment I've ever received & couldn't agree more with everything you've said. Thanks for taking the time to write this out. Too bad you've lost most of the data comparisons you've made over the years, it sounds like it would be a wealth of knowledge that would be invaluable to the younger generation of surveyors coming up!
Amazing , thank you so much for the video and very useful information.
Thank you for this video. I've always been curious about how accurate, and precise, these measurements may be. Great job and i plan to watch your other videos
Great video and appreciate you simplifying it for us!
Look forward to your future comparisons video
Excellent video !
Thank you for taking the time to share.
Great video👍
Good video, very informative.
Thanks, this video will save me hours of explanation ^_^
This is fantastic work sir, hats off !
I would be interested to understand the part of GNSS receivers spec talking about RTK. Specs usually says 8mm + 1ppm for rtk (assuming nearest base in corrections network) and 8mm +0.5ppm for network rtk. But 0.5 ppm of what ? Distance to a closest base ?
Distance from whatever base you are receiving corrections from. If you're talking about a network solution the virtual base is right beside you, theoretically removing that portion of the accuracy spec. That being said I have been told from a fairly reliable source that single baseline is preferred when there is a base station within 9 miles, beyond that a network solution is supposed to be better. That being said I haven't personally tested this theory so cant say for sure if it's right or wrong.@@milannovak4772
@@The3rdDimensionSurveying As I understand it, in a network solution, the vectors you get are from a physical reference station. So yes, the PPM error does still apply, typically reported as 0.5 ppm for network solutions. The virtual reference station (VRS) is merely a point where RTK corrections are computed. That is, the network detects your rover autonomous position upon initialization and interpolates correction values for that location. That RTK correction message is considered valid within a 5 kilometer radius circle (per a Trimble expert I met in 2017). However, every time we lose initialization, the VRS point resets. Careful inspection of your RTK vector raw data should show the physical reference station name. If you'd like to discuss it in more detail, please reach out.
Hello my friend! Thanks for the aweosme videi! Do you have any video telling us about working with total stations and GNSS?
Wich method is the best to get all data fitted: site calibration, applying avarage combined scale factors for the entire site or applying combined scale factor for every single point of your project, accordingly to the N,E and ortho height of the point? How to survey out of site calibration and still get accurated survey data, like the total station measurements? For example for larger sites or linear projects, which method would you use to convert the GNSS data to ground? I struggle so much with these questions and how to keep things under control. Thanks if you could bring some light on this topic!
Hello! Hands down using STAR*NET. 100% the right tool for that task. I don't have a video on exactly how to do that yet, but it is on the To Do list. Hopefully within the next month or two. I really makes it amazingly easy to switch between grid and ground by exporting two separate control files and automatically calculates your CSF based on the point you want to scale about. Its fantastic at merging total station and GPS data in the same project as well. There may be other Least Squares Adjustment programs out there that do the task as well but all the others I've tried don't hold a candle to STAR*NET.
Great video, thank you. The link to your datsheet seems to not be working, is it possible fix this?
Great video! I think most surveyors I know have pushed the capabilities of GPS in sketchy locations.
Are you aware of any manufacturers that have atmospheric inputs at base set up. That could be a great software feature. I always input them on TS set ups.
Unfortunately inputting surface pressure and temperature wouldn't do you much good. It's not quite as straight forward a calculation such as the one for your EDM on your total station. There is real time modelling already being applied for Troposphere delay however. The error left over from troposphere delay is from imperfect modelling.
I can sort of see the manufacturers point, you could use their product with many different correction sources all with different outcomes in precision...
I have a question. If the bestestation requires very accurate absolute position to calculate corrections from measurement, and be close to the receiver (probably less than few km for decent results), but is mobile, how its gets its own accurate position in the first place during the setup? I understand if this is permanently on a building, we can measure positions for a month, and got good accuracy, using multi band receiver, averaging a lot of errors, and taking favorable conditions. (reference stations use more direct methods, including laser ranging, but these methods are not available for normal use). But if the base station is mobile how this can be done? I think the answer is simple. It is not possible. the only thing it can do is improve precision, but not accuracy, at least not without very extensive averaging of the base station position. In many scenarios high precision and medium accuracy is all you need, including survaying and geodetics, especially for construction sites. Especially if there are some other datums we are using as a reference point. But it is not suitable for all applications.
I think your question is how do we get an accurate absolute position of the base station? One option is to collect and process static data. When you say it's not possible I guess it comes down to your definition of an accurate position.
It's not true we cannot improve absolute accuracy of the base station, we can take repeated observations. Just like any measurement that has random error, the more blunder free observations we take, the better the averaged result will be.
Really great video. Do know a free software that can help convert grid coordinates(RTK) to ground coordinates (Total Station). I know TBC can do this but I am hoping that there's some other low cost or free software that can do same. I have been searching for this for a while.
Sure, it can be done with nothing more that a service that gives you a combined scale factor for one of your points on site, NCAT for example (www.ngs.noaa.gov/NCAT/), then scale your points by the inverse of the CSF. Which can be done in CAD, excel, or even a calculator if you want. But I would be very careful when doing this, some of the biggest blunders I've ever seen were from improper handling when going from grid to ground. I'm talking millions of dollars to rectify mistakes caused by this.
@@The3rdDimensionSurveying Thanks a lot. I wish there was a video to guide me but I will try it and see.
@@geossotechltd It's on the list! Good Luck!
@@The3rdDimensionSurveying Just discovered its mainly for America. I can't find my regions input reference frame
Qgis can do this
please reupload datasheet this item might not exist
1drv.ms/x/s!AkXoacH8d5pAhJoQKFZxGUBbTAsqzA?e=5DT3qm
@@The3rdDimensionSurveying THANKS
It’s not GPS. It’s GNSS + RTK.