Simple radio setup to mark property lines..?

[IdahoMan]

Hello,

Been thinking about how one could mark (flags, posts, whatever) property lines through hilly terrain or dense forest where line-of-site isn't possible. The property corners are known and marked, but you cannot see from one to the other.

Can a simple setup be done where a transmitter broadcasting a simple tone is placed on one corner, and some kind of directional receiver be placed on the other..? The main goal is to have the receiver antenna be pointed directly (+/- 1ft) at the broadcasting antenna.

No doubt they have equipment that can do this, but it'll cost a ridiculous amount. Besides, something like this has got to be really easy to make with junk parts, it's just getting the right antenna setup that must be the important part for the above stated accuracy.

Any recommendations? Point me in the right direction (no pun intended).


Thanks,
IM

[Ian.M]

No.  For anything reasonably accurate, you'd need to use very large parabolic antennae and a frequency up in the millimeter wave band, and that's strictly line of sight stuff.

If there's a reasonable sky view, a survey grade GPS could help, but GPS systems are problematic under dense forest canopy.

[janoc]

If there isn't line of sight then directional radio will not work neither. You would need microwave frequencies to achieve reasonable accuracy and those get blocked the same as light does.

Why not simply use a GPS receiver and "walk" the property line based on a map? That's probably the most accurate method (+- 10m or so with a normal consumer grade receiver, down to centimeters with a special surveying GPS capable of receiving the differential correction signals) available unless you call an actual surveyor. Anything radio is not going be any more accurate than this.

A surveyor will likely split the property into pieces where there are lines of sight and can use normal optical equipment. Then they will determine the property line in a piecemeal fashion, triangulating from one piece of the property to the next.

[rhb]

Look into PPP aka Precise Point Positioning GPS.

Here's something I was sent.  I've looked at it briefly, but have not done a thorough study.  The alternative is differential GPS using an Arduino, GSM module and GPS.

Differential GPS is what the surveyors use.   It's about US$100 for the hardware plus the cost of two cell lines.   For relatively short distances, one of several RF chips should work.  The more difficult problem is converting a metes and bounds description into lat-lon.  I've done this in the context of processing deviated oil well data on a Unix workstation, but not in the context of an MCU.  Probably the thing to do would be to convert to lat-lon on a PC, load the lat-lon waypoints and then do crude approximations to get you to the corners.

RTKLib is a software toolkit by Tomoji Takasu for working with raw data capable GPS hardware. (Like much Ublox hardware although recently they have been making this harder in their less expensive chips.)

Let me give you some URLs..

http://gpspp.sakura.ne.jp/indexe.html  (Start here)
http://rtklib.com/   (and here)
https://github.com/tomojitakasu/RTKLIB
http://wiki.openstreetmap.org/wiki/UbloxRAW  (this focuses on Ublox, other manufacturers HW may also work..)
http://wiki.openstreetmap.org/wiki/RTKLIB  (list of compatible hardware but is missing some)
https://rtklibexplorer.wordpress.com/2016/02/03/collecting-raw-gps-data-with-rtklib/

I am also quite interested in being able to locate a single corner and then find the others.  I've got all or most of the hardware, but haven't gotten down to work yet.

[IdahoMan]

Hhhmm..

Just how accurate can a recieving setup be on focusing on a signal source, depending on the frequency?

I know if I where to light a lightbulb, place a photosensor in a box with a deep slit (say 1" deep and 1/8" wide) on a side of the box, and rotate the box I could get it pointed at the bulb.

The frequencies that can travel over land without LOS, can they not be tracked down to pin-point accuracy due distortions and reflections? (I'd love to see the math on some of this. v=fw, etc.)

I love radio/EMR, it's a fun and empowering subject.

[janoc]

The frequencies that can travel over land without LOS, can they not be tracked down to pin-point accuracy due distortions and reflections? (I'd love to see the math on some of this. v=fw, etc.)

I love radio/EMR, it's a fun and empowering subject.

If the frequency can travel over land without LOS, its wavelength has to be significantly longer than the size of the obstacles so that it can bend over them instead of being reflected or blocked.

The consequence is that such localization is going to be terribly inaccurate - you can localize e.g. an AM broadcast station by triangulating it (take a directional antenna and determine the bearing to the station from at least two places). The accuracy will depend on how far apart those places are - the farther apart, the more accurate it will be. Furthermore, those places need to be far enough from the transmitter (many kilometers), if you are too close you won't be able to take a sufficiently accurate measurement because the signal will be very strong even when pointing the antenna in the "wrong" direction.

Anyhow, with this approach you won't get anything more accurate than perhaps few kilometers, depending on the wavelength of the signal. If the signal can bend over an obstacle vertically it can do so in the horizontal plane too - screwing up the accuracy of your angular measurement. Good enough to roughly localize a rogue station in across the globe in Russia but useless to survey a boundary of the property.

With high frequencies you can do better but then those will not refract around obstacles and will get blocked or, worse, reflected and/or interfering, giving you false signals.

[MasterT]

Hhhmm..

Just how accurate can a recieving setup be on focusing on a signal source, depending on the frequency?

To make it compact & portable, antenna size has to be small, and consequently frequency pretty high. I did ones direction finder for 433MHz band, using two antenna and phase calculation done by arduino.  Accuracy was about 1 degree.

[janoc]

Here is a good description of how these systems work:

https://shorties.be/podcast/Direction%20finding.pdf

[rhb]

Raydist, which was widely used in the early 80's, would give you 20-50 feet (*if* you had a good base station setup).  Without that or with older technology, 100-300 feet was the best you could do.  And this was offshore, cost is no object, seismic data acquisition.

Basic GPS is much better than Raydist.  Look at Takasu's work.   He can solve raw GPS to centimeters from what I've read.  These are professional papers, so it's not BS.  He's a bit cagey about the fine print, but he'll let you use his software for free.

Otherwise, differential GPS with a fixed station at one corner will give centimeter accuracy for  under $300 in hardware. GPS errors are dominated by variations in moisture in the air. Differential GPS takes this into account over small distances.

EDIT:

All radio positioning systems rely on timing differences among multiple stations.  Directional sensing it not usable for anything other than locating a transmitter very approximately.

For a wooded property of any size, an optical survey is a *lot* of work.  I surveyed 40 acres on a steep mountainside as a teenager.  It was just me and my Dad, so I got to do all the line clearing and the tail end of the chain. It took several weeks to do.

You can do post processing differential GPS with ordinary GPS units.  At one of the corners record the location at 1 minute intervals.  Use another GPS to walk out the line,  flagging and recording at the same times.  You can then go back, calculate the corrections for each station and move the flags.  You'll need to project the locations to state plane coordinates to determine the distance and direction for each correction.

An electronic property line would require a buried wire and a receiver a la the electronic dog fences.

[DrGeoff]

A laser and a series of lenses/mirrors along the way?

[janoc]

A laser and a series of lenses/mirrors along the way?

How does a laser get through OP's dense forest? 

[ahbushnell]

A laser and a series of lenses/mirrors along the way?

How does a laser get through OP's dense forest? 

It burns it's way through.   

[Mjolinor]

Gese technology for technology's sake.

Calm day, two balloons and two long bits of string. Assuming it is a straight boundary.

[IdahoMan]

Gese technology for technology's sake.

Calm day, two balloons and two long bits of string. Assuming it is a straight boundary.

That was the next thing I was thinking of.   It would have to be a calm day though for sure. Then use a rifle scope and laser projector (like a grocery store scanner has) to paste a line. The scope to aim at the the balloon string, and the laser to project a line.

[Mjolinor]

Thinking about this you don't need a calm day really. The two strings will only align all the way down if there is no wind. If there is wind then they will form a V if you are on the boundary line, the bottom of the V being the boundary. If my spatial visualisation is correct that is.

[hendorog]

Locate a transmitter at each corner and measure the phase difference between the recieved signals at different locations. Then trigonometry.

[cdev]

The cost of RTK-capable GPS receivers are very low now. You can even buy multi-constellation-capable GPS modules made for UAVs for around $20 (which will require a USB-UART converter - another $1.50-4 and a USB-capable computer with some kind of network connection to work.  It could also be done with an RTK-capable smartphone. See rtklib.com for a list of compatible HW. A lot of ublox hardware sold on ebay for drone use are variants on the ublox m8n that use the older 2.02 firmware (and cannot be upgraded) that can be used with RTKlib. the newer firmware likely needs some less well documented commands that people don't know yet. Galileo compatible M8Ns running 3.x firmware are probably NOT compatible with RTKlib unless somebody has figured out the new commands by now.  So then one has to spend around $70 to get the m8t which does support RTKlib and adjustable timing pulse as well.  The Navspark NS-hp does support single constellation RTK and its very accurate but two or more constellations probably impart big advantages under heavy tree cover.

Look into PPP aka Precise Point Positioning GPS.

Here's something I was sent.  I've looked at it briefly, but have not done a thorough study.  The alternative is differential GPS using an Arduino, GSM module and GPS.

I am the person who made the suggestion below, which is I think the best option you have. Normally the accuracy of RTK (using a base station) is superior to PPP. There is no reason you cannot use post processing - Try to get the best possible line of markers using PPP as described on his web site and also on rtkexplorer. Then attempt to improve the accuracy using post processing of the data collected then using data collected from a local base station (the nearest ones you can find) at the same time. Its quite possible that there is free base station data for your area.

There is a national CORS net of RTK base stations, and many states are part of it. Where I live both the state I live in and a neighboring state both have free RTK data available for everybody if you sign up for an account. Try using it to measure other things first spend a bit of time learning how to do it. The rtkexplorer web site is a good resource but you should also check out the authors site rtklib.com and also the other links I included below.

RTKLib is a software toolkit by Tomoji Takasu for working with raw data capable GPS hardware. (Like much Ublox hardware although recently they have been making this harder in their less expensive chips.)

Let me give you some URLs..

http://gpspp.sakura.ne.jp/indexe.html  (Start here)
http://rtklib.com/   (and here)
https://github.com/tomojitakasu/RTKLIB
http://wiki.openstreetmap.org/wiki/UbloxRAW  (this focuses on Ublox, other manufacturers HW may also work..)
http://wiki.openstreetmap.org/wiki/RTKLIB  (list of compatible hardware but is missing some)
https://rtklibexplorer.wordpress.com/2016/02/03/collecting-raw-gps-data-with-rtklib/

PPP takes a long time to settle down under trees.

have you tried a good cheap multiconstellation GPS? If you live on the property, you might be surprised at how accurate a cheap two constellation GPS might be just running on its own in pedestrian or better yet, stationary mode. (thats telling it its not going to be moving while you are measuring)

the Navspark-mini deal will only cost you $10. You'll need a laptop. The (non-rtk) accuracy of the navspark mini when at least one beidou sat is visible is *extremely good* for a single constellation GPS that costs under $10.  Also some of the Ublox GOS's *after they have been running in the same location for a while and with the Kalman filter set properly* are accurate to under 2 meters fairly consistently, once they settle down. You have to let them run for awhile. Also use a good antenna, with a good ground plane under it.

[cdev]

This is not as difficult as one might think it would be.


First and most simple approach, first determine the exact heading by creating two placemarks in Google Earth and determining the heading of the angle between them.

You could then generate a path in KML format of the path - that will be the border.. Keep that handy..
there are several things you could do.. Some dont even require a GPS.. Simplest would be to set off on foot following the heading until you get to the other end, either unrolling a string or tying markers to trees or both..

Using a good compass, keep going until you find a tree at exactly the right spot, - continue marking trees with colored pieces of cloth in a high visibility color either when you hit a tree at the exact spot or when you're about to lose visibility make some kind of marker using a stick and cloth, keep going until you get to the end.

As somebody who has done this a few times, I suspect that as you are doing this you will likely find markers from previous surveys - even if they were done some time ago, once you are there and your exact location is informed by the other methods.


 When you get to the other marker, if the path determination has been fairly successful you'll be able to tell by how close you are at the end. Then go back and check your work with a GPS and if it seems appropriate, improve the marker locations.

Really dense trees make it hard to get it right but keep in mind that if you have a good GPS which doesnt need to be at all expensive.. if the GPS is configured correctly, the longer you let the GPS sit still, the better the measurement will become..

If you take your time, you would be surprised how well you could do under slightly less dense trees with some warm up time (a few hours, maybe as you are doing other stuff elsewhere, just leave your GPSs both on and connected to antennas).


If you use RTK, you'll likely with some practice be able to get much more accuracy, (just a few centimeters - without a fancy surveying grade GPS antenna) depending on how thick the trees are and how long you are willing to wait to have the solution home in on the right number..

Try some of the links you get back from this search and the other URLS

https://www.google.com/search?hl=en&as_q=idaho+cors+network

Hello,

Been thinking about how one could mark (flags, posts, whatever) property lines through hilly terrain or dense forest where line-of-site isn't possible. The property corners are known and marked, but you cannot see from one to the other.
....

No doubt they have equipment that can do this, but it'll cost a ridiculous amount. Besides, something like this has got to be really easy to make with junk parts, it's just getting the right antenna setup that must be the important part for the above stated accuracy.

Any recommendations? Point me in the right direction (no pun intended).


Thanks,
IM

You may not even need centimeter level accuracy (thats what you get from RTK.)
Newer Multi-constellation consumer grade GPS hardware IF CONFIGURED PROPERLY and given a few hours to warm up (it can be traveling that time) can be a lot more accurate than many people think it is. In the open, within just a few meters or even better.

You say you know the exact position of the markers? If so, you can leverage that info to make your relative measurements much more accurate. get two identical cheap GPSs and as somebody else said, keep one at your known reference point and bring the other one with you. Record the data from both the stationary location and the moving one.

You then can diff the movements in the known stationary receiver's track from the other track too to get substantially better accuracy.

You would be surprised how good modern GPSs are.. even without RTK or DGPS.. read this article about the improvements from adding additional GNSS systems.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4321187/

I'm kind of distracted right now by a totally unrelated issue so this and my other posts may seem a little scattered but I would be happy to walk you through this try to answer your questions or point you to where you could get expert help (much more so than me!)

[cdev]

A laser and a series of lenses/mirrors along the way?

How does a laser get through OP's dense forest? 

mark the tree that blocks it with a brightly colored circular marker that goes around it completely and mark the point on the other side that seems to be the continuation of the line from the other side with a brightly colored pen or reflective marker tape??

[tautech]

I helped a mate find some # of missing survey pegs on his mainly bush covered block a few years back.
There are a few tricks you can use especially if the original survey was done before GPS days.

Optical LOS was used back then and not along the boundary line but adjacent to it.

You'll need some HW to do this but most importantly a lodged copy of the original survey plan.
In it there's a # of lettered codes you need to decypher, IS=Iron Spike, IT= Iron Tube/Pipe and when they've used old existing reference points O is added to the coding.
The spikes and tubes are where the theodolite has been positioned and bearings and measurements to each element of the survey are taken. Understand this is where things get interesting as all measurements taken, regardless of grade are converted to horizontal/level for the lodged plan and become shorter on the level plane that drawings are based on.

Anyway, armed with that basic knowledge you'll need some sort of metal detector to find these spikes and tubes, and a long tape for measurements but the most critical is a optical theodolite to set the LOS angles between points.

The old surveyors belayed along survey lines whether they be LOS or an obscured path and always had 2 reference points in view.
My buddy is an excellent mathematician however the calculations were daunting to find the actual boundary line and how it related to the buried spikes/tubes in the ground as they don't, they are only an accurate belay to an end point/peg. As each boundary peg is only related to belay points and horizontal measurements (and between them calculated) the maths gets very involved quick.
Nonetheless after a few sessions we found them all in heavy cover where the original LOS paths were now overgrown.

BUT we still employed an old trick my dad taught me when trying to find the path a fence would take in broken ground. Two long stakes, where you sight over them backwards to a reference and forwards to as far as was practical where another stake was placed to start the process over.
Using this method, calcs and measurement along with traditional optical survey methods we found all the tubes/spikes and pegs and marked them all for prosperity with steel standards. 

The most useful tool was a surveyor's metal detector like this:

Exceptional sensitivity and directivity.
 

[DrGeoff]

A laser and a series of lenses/mirrors along the way?

How does a laser get through OP's dense forest? 

That's what the mirrors are for. 

[janoc]

A laser and a series of lenses/mirrors along the way?

How does a laser get through OP's dense forest? 

That's what the mirrors are for. 

And how many of those are you going to need along a line through a forest? All the while keeping them perfectly aligned all the time while you are surveying? A minuscule movement of the mirror (e.g. due to wind swaying the tree to which the mirror is affixed) means a huge error at the far end.

This is not a practical method at all.

Furthermore, if the forest is sufficiently open to use a laser (even piece-wise) then you can use standard surveying equipment (theodolite + reflectors) and triangulation as well.