Innovative Scientific Uses of GPS

[anorlunda]

TL;DR Summary

New ways to use GPS for scientific purposes.

https://arstechnica.com/science/2019/12/gps-is-going-places/

I really liked this article. It is not too long. Read it. The 5 ideas are:

1. Feel an earthquake (@davenn will like this)
2. Monitor a volcano
3. Probe the snow
4. Sense a sinking
5. Analyze the atmosphere

 

[Dullard]

I love the last sentence of the article:
"Not bad for something that was just supposed to help you find your way across town ."

That isn't why the Air Force funded GPS. GPS was (primarily) for guiding weapons and (arguably) as a first-strike forward observer. It was initially operated in a mode that actually denied precise navigation to unauthorized users. That went away during Desert Storm when it was easier for a platoon commander to get a commercial receiver shipped from his Mom than to get an 'authorized' receiver from his Uncle Sam. Mr. Clinton made it permanent. It's almost an accident that we have access to this incredibly useful tool.

 

[FactChecker]

I love the last sentence of the article:
"Not bad for something that was just supposed to help you find your way across town ."

That isn't why the Air Force funded GPS. GPS was (primarily) for guiding weapons and (arguably) as a first-strike forward observer. It was initially operated in a mode that actually denied precise navigation to unauthorized users. That went away during Desert Storm when it was easier for a platoon commander to get a commercial receiver shipped from his Mom than to get an 'authorized' receiver from his Uncle Sam. Mr. Clinton made it permanent. It's almost an accident that we have access to this incredibly useful tool.

There are a lot of examples of advanced technology that was developed with military uses in mind, but were eventually of great benefit to the general public. That is where a lot of your tax dollars go and the results are predictable. That primary funding for things like integrated circuits, the internet, jets, etc. were military.

PS. A smartphone can use both the GPS and the cell phone tower information to get more accurate location results than plain satelite-based GPS.

 

[davenn]

Summary:: New ways to use GPS for scientific purposes.

1. Feel an earthquake (@davenn will like this)

Thanks for the heads-up

It's more about ...

—to track geologic processes that happen on much slower scales, such as the rate at which Earth’s great crustal plates grind past one another in the process known as plate tectonics. So GPS might tell scientists the speed at which the opposite sides of the San Andreas Fault are creeping past each other, while seismometers measure the ground shaking when that California fault ruptures in a quake.

This has been getting done since the early 1990's. I remember when I doing geology at university back
then, one of the lecturers giving a public talk on how this was being implemented. At the time, for most of
us, this was WOW and COOL. And now, over the last 25 years, this has become just a common thing to do.
GPS quality has been improved and with many more satellites up there now ( not even mentioned in that
article) There are a number of constellations. The two major ones -- GPS - USA, GLONASS - Russia, Now
more recently - BeiDou - China and Galileo - Europe. And a couple of other countries with just a few sat's
up there - IRNSS - India and QZSS - Japan.
Who knew that 20 years later, 2011, I would start working in this field of high precision GPS.

A few basic facts, car/handheld GPS units - ~ 5 - 10m accuracy. Even our high precision units, when
used on their own, are only ~ 1 - 5m accuracy. This precision gear is never used on its own, rather it is
connected to a local GPS base station by either radio or internet, where it can receive GPS corrections
and get much higher accuracy. Originally Differential GPS was used and it improved accuracies from
several metres down to several cm. This was further improved by incorporating RTK positioning.
This improved accuracy even more and it is now easy to get down to 2-3mm horizontal positioning.
Vertical positioning is usually always around 2 to 3 times less accurate than horizontal eg. 2-3mm Hz
would give around 5 - 7mm vertical accuracy.

Now, to put this into use, when studying geological plate tectonics. Take a plate boundary like the
San Andreas Fault in California, USA or the Alpine Fault in the South Island of New Zealand.
It is quite easy to measure the 2 significant motions along that boundary even before an earthquake
occurs. And after a quake occurs, then very accurate measurements of the motion can be measured.

Lets look at a drawing as well as an actual fault.

In the above drawing ( a represation of the Alpine Fault in NZ) I have shown relative ground motions
close to and away from the faultline. This fault isn't pure strike-slip, rather there is a component of
compression and so it is called an oblique-slip fault.
At the fault, there is no motion at all and as you move away from the fault axis, the relative ground
motion continues as indicated by the size of the arrows. When the fault finally ruptures, all that motion
that has been occurring away from the fault then releases at the fault and it "catches up".

The ongoing ground motion at position 4 is much more substantial than at position 3 or 2. And for the
Alpine Fault, the motion is ~ 40mm / year. This motion is far above the 2 - 3 mm accuracy noise of the
GPS system so is easily measured and graphed.

The above image shows just a small, ~ 150 - 200km section of the Alpine Fault. It can be seen as that
line in the topography running from lower left to upper rightOK I will leave it at that before I hijack the thread too muchDave

 

[anorlunda]

I love the last sentence of the article:

I also liked the description of the author.

Alexandra Witze is a Colorado-based freelance journalist whose work has appeared in Nature, Science News and other publications. She does not have a good sense of direction.

Originally Differential GPS was used and it improved accuracies from
several metres down to several cm. This was further improved by incorporating RTK positioning. This improved accuracy even more and it is now easy to get down to 2-3mm horizontal positioning. Vertical positioning is usually always around 2 to 3 times less accurate than horizontal eg. 2-3mm Hz would give around 5 - 7mm vertical accuracy.

I read about using DGPS used to sense wingtip flexing on airplanes. Absolute accuracy would not be as important as the ability to detect incremental changes. It is non-contact distance sensing as opposed to non-contact position sensing.

 

[FactChecker]

I read about using DGPS used to sense wingtip flexing on airplanes. Absolute accuracy would not be as important as the ability to detect incremental changes. It is non-contact distance sensing as opposed to non-contact position sensing.

DGPS is really very widely used. As long as one position is very accurately known (e.g. survey markers), using it for offset positions in DGPS is possible. It's just not what is used by the general public in common devices.

 

[Tom.G]

At least for a while, DGPS was used by Golfers. A local broadcast radio station was 'convinced' (probably paid) to send correction data on a sub-carrier. Of course the absolute position of the station was known. The Golfers had receivers that took the received GPS position and applied the local correction. They then knew exactly how far they were from the hole they were shooting for. This was circa 1980 + or -.

 

[Fred Wright]

Ten years ago I was employed by a firm which used gps to track wildlife. A collar containing a device with a micro-controller would integrate a gps receiver and a uhf transmitter and relay the position data to the ARGOS satellite system. I was tasked with developing the firmware for a collar to be attached to marine mammals. The overriding issue was to obtain a valid gps fix in less than three seconds from a cold start(i.e. the animal only surfaces for three seconds). By incorporating a clever algorithm, using ephemeris data, we succeeded in getting the "quick fix" and the device was shipped to several oceanographic institutes.

About a year ago some Swedish fishermen found a Beluga whale entangled in their net with a collar containing a mysterious device. They brought it to the local authorities who could not make heads nor tails of it. The news media picked up the story and posited the theory, confirmed by some Swedish politicians, that the Russians were enlisting Beluga whales to spy on Sweden. This, despite the fact that the collar was clearly marked, in the English language, "PROPERTY OF FIO" (Florida Institute of Oceanography).

 

[phinds]

The news media picked up the story and posited the theory, confirmed by some Swedish politicians, that the Russians were enlisting Beluga whales to spy on Sweden. This, despite the fact that the collar was clearly marked, in the English language, "PROPERTY OF FIO" (Florida Institute of Oceanography).

SO ... the Russians infiltrated the Florida Institute of Oceanography

 

[anorlunda]

that the Russians were enlisting Beluga whales to spy on Sweden.

Uh Oh. They will be depth bombing them next. :-)

 

[mfb]

A real-life example of GPS measurements of an earthquake: https://www.researchgate.net/figure/Monitoring-of-the-PolaRx2e-GPS-antenna-position-at-LNGS-showing-the-slow-earth-crust_fig6_278812102 - an 800 km distance measured with millimeter accuracy.

 

[Swamp Thing]

SO ... the Russians infiltrated the Florida Institute of Oceanography

Seeing that FactChecker has liked this post, I take it that it's a fact.

 

[FactChecker]

Seeing that FactChecker has liked this post, I take it that it's a fact.

Please excuse my choice of nickname. I used it on another web-site in another context and I just reused it without thinking. I have been stuck with it since then. It's embarrassing to me on this forum. :>(

 

[phinds]

Please excuse my choice of nickname. I used it on another web-site in another context and I just reused it without thinking. I have been stuck with it since then. It's embarrassing to me on this forum. :>(

Well, just get everything right and it won't be

 

[FactChecker]

Well, just get everything right and it won't be

That ship sailed long ago. :>)

 

[jedishrfu]

We could always change your name to @Snopes and make you a StarWars character that gets murdered by an over zealous student. Oh wait they made that movie already with @Snoke better stick with @FactChecker. I think its a cool choice for a name.

 

[Svein]

About 15 years ago an application was proposed - use GPS to detect whether or not a football had actually passed the "goal line". I wrote a lengthy report and proposed a sort of "backward" GPS solution, but questioned whether or not the players would accept a football with electronics inside (it would probably alter the spin properties of the ball).

 

[jedishrfu]

They have tech to do ball out of bounds checks but it doesn’t use internal ball electronics. It uses video analysis Via strategically placed cameras.

https://en.wikipedia.org/wiki/Electronic_line_judge

i wonder if an rfid chip in the ball could work And what issues you might have. The line would need electronic sensors too.

 

[Svein]

i wonder if an rfid chip in the ball could work And what issues you might have. The line would need electronic sensors too

That was my question at the time too, but the customer liked the GPS version. And the customer is always right?

If I remember correctly, the customer wanted to patent my solution, but if so, they did it without involving me.

 

[anorlunda]

I too think that a video would be more successful. It is easy to embed a video camera in the goal post. It is easy to feed the recorded video to screens that the referees, fans, and players can see.

Edit: On the other hand, using GPS for a sport like orienteering would be a very good match.

 

[jedishrfu]

That was my question at the time too, but the customer liked the GPS version. And the customer is always right?

If I remember correctly, the customer wanted to patent my solution, but if so, they did it without involving me.

Thats just wrong. You should be on the patent and if they wanted it should get you to transfer the rights to them.

At my company, they claimed everything but would compensate you for your patentable ideas.

 

[FactChecker]

Thats just wrong. You should be on the patent and if they wanted it should get you to transfer the rights to them.

At my company, they claimed everything but would compensate you for your patentable ideas.

Unless he had taken action to legally protect his rights before he gave them the information, his rights are gone.

 

[Svein]

Thats just wrong. You should be on the patent and if they wanted it should get you to transfer the rights to them.

As I said - I don't know. I have several patents in my name, but assigned to the company I was working for at the time. Patents just meant a tough job in front - checking out possible infringements and writing the patent claims as watertight as possible. And before you suggest using professionals - the last time I did it, I did not recognize the results and wondered if I had sent in the wrong application!

 

[jedishrfu]

Yes, my company’s IP department handled the patent details and it was amazing with what they came up with as examples and how it worked. The key is not the description but the claims made. Most court cases debate the claims to decide if there’s been an infringement.

This American Life did a great podcast and follow up podcast called When Patents Attack that you should listen to. In one case, a single apostrophe decided the fate of the patent and saved the day for one company thought to be infringing on this overly broad patent used successfully against many other companies who decided to just pay up and not fight it in court.

 

[jedishrfu]

Unless he had taken action to legally protect his rights before he gave them the information, his rights are gone.

Sometimes you can’t or your company is just not interested in doing so. In this case, it could be the client insisted on GPS and he became the implementor of the idea.

 

[Swamp Thing]

About 15 years ago an application was proposed - use GPS to detect whether or not a football had actually passed the "goal line". I wrote a lengthy report and proposed a sort of "backward" GPS solution, but questioned whether or not the players would accept a football with electronics inside (it would probably alter the spin properties of the ball).

Did your solution involve a local augmentation transmitter of some sort?
I remember reading that raw GPS data is actually quite noisy, and they do some Kalman filtering to extract the maximum-likelihood position and velocity from the noisy data. But when you kick a football , your Kalman filter has to forget the old information and find the new velocity and position from perhaps a couple of data samples. How would that work in practice?

 

[jedishrfu]

Not to mention that the US Government limited accuracy for a time to within +-5m but since 2000 its now at +-30cm.

https://en.wikipedia.org/wiki/Global_Positioning_System

 

[Svein]

Did your solution involve a local augmentation transmitter of some sort?
I remember reading that raw GPS data is actually quite noisy, and they do some Kalman filtering to extract the maximum-likelihood position and velocity from the noisy data. But when you kick a football , your Kalman filter has to forget the old information and find the new velocity and position from perhaps a couple of data samples. How would that work in practice?

As I said - it was a "backward" proposal, involving a transmitter in the ball and several receivers along the edge of the stadium. The receivers would synchronize their clocks the same way the GPS satellites synchronize their clocks.

The reason we discarded the pure GPS solution was due to some basic measurements. I put two GPS receivers next to each other on the roof and read the position information from both over a couple of hours. The position difference (which should have been constant) fluctuated constantly.

 

[mfb]

Not to mention that the US Government limited accuracy for a time to within +-5m but since 2000 its now at +-30cm.

https://en.wikipedia.org/wiki/Global_Positioning_System

That's the absolute position uncertainty. If you take differences between receivers nearby you get a much better resolution. Differential GPS can reach millimeter uncertainties.

 

[davenn]

Differential GPS can reach millimeter uncertainties.

Yup, that's where RTK positioning really comes to the fore as I commented on back in post
#4. Differential is worse than RTK as it can only do a couple of cm, RTK gets down to the mm accuracies

Originally Differential GPS was used and it improved accuracies from
several metres down to several cm. This was further improved by incorporating RTK positioning.
This improved accuracy even more and it is now easy to get down to 2-3mm horizontal positioning.
Vertical positioning is usually always around 2 to 3 times less accurate than horizontal