Hur fungerar ett korrektionsnätverk? Hur fungerar spoofing och jamming? Solstormar? Äldre och nya GNSS mottagare, vad är skillnader?

Mät på engelska!

Links:

• Gavin Schrock, Author at xyHt
• Gavin Schrock | LinkedIn
• GPS World
• Gavin Schrock – GoGeomatics
• GPSJam GPS/GNSS Interference Map

Varmt välkomna ska ni vara till ännu ett avsnitt av Mätpodden den här sommaren 2023. Det här är ett fördjupningsavsnitt med Gavin Schrock, en så kallad mätguru, en väldigt erfaren herre från USA som jag intervjuade i förra avsnittet om GPS och solstormar och så vidare. Gavin upptäckte jag för att han skriver väldigt mycket på webben. Om man läser xyHt magazine, Geomatics eller GPS World så skriver han många artiklar där, och väldigt bra artiklar som har en väldigt bra blandning mellan teori och praktik. Det var så jag fick upp ögonen för honom. Jag intervjuade honom, och mycket av det vi pratade om kom inte med i det avsnittet om GPS, även om en del gjorde det. Så jag tycker att det är värt att lyssna på hela det här, där vi faktiskt pratar om hans bakgrund men också om hur han driver ett korrektionsnätverk i USA.

Han drar också en del jämförelser och lärdomar från SWEPOS, som han har besökt i Sverige. Självklart pratar han om det ni kanske redan har hört, om störning och utslagning av GNSS-signaler, alltså spoofing, jamming och solstormar. Men vi pratar också om GPS. Vad är skillnaden mellan en dyr och en billig mottagare? Han har testat väldigt mycket. Hur fungerar det att mäta under träd? Och lite om framtiden. Vart är utrustningen på väg enligt hans erfarenhet? Han ger liksom fem till tio minuter långa svar på frågorna som om de vore väl förberedda, men han är helt oförberedd på frågorna. Verkligen en väldigt intressant herre. Så lyssna gärna på det här och gå in och läs artiklarna. Det finns mycket att lära, både om LiDAR, drönare, GPS och egentligen både det ena och det andra i de här artiklarna.

Jag kommer att länka till de ställen där han skriver. Och innan jag drar igång det här avsnittet vill jag ändå göra den första riktiga reklamen för mig själv här. Gå in på landskaparen.se, som nu lanseras som någon form av betaversion. Där kan man faktiskt boka den här kursen som jag har pratat om. Jag har bestämt mig för att fokusera på anläggaren. Så om du identifierar dig som anläggare, alltså du kanske är i mätbranschen, men jag vet att det är mycket folk från anläggningsbranschen som lyssnar på det här. Om du jobbar i anläggningsbranschen, oavsett om du är maskinist eller arbetsledare, du kanske är mätare eller någon form av projektör, men om du har en koppling till anläggningsbranschen och vill förstå koordinatsystem, förstå hur maskinmodeller fungerar och kunna skapa dem, och vill veta hur DWG:er fungerar. Och egentligen vill du förstå det här innan du har tröttnat på att vänta på resultatet.

Du känner att: det här borde jag kunna göra själv, åtminstone de enkla delarna. Du kanske känner att du behöver betala mycket för ett resultat som inte motsvarar dina förväntningar. Du känner att: jag vet hur jag vill lägga den där planen eller det här schaktet, men jag kan inte uttrycka det själv i en modell. Ja, då ska jag lära dig det. Och det börjar med steg ett, och jag kallar det den enkla villan. Där lär du dig de här grunderna. Jag har under åren skapat det som ett ramverk, alltså en process. Det är den jag kommer att lära ut, oavsett om det är det första projektet, den enkla villan, eller som det kommer senare, en industritomt eller den komplicerade byggarbetsplatsen. Det du alltid behöver göra är att ta reda på förutsättningarna. Det betyder koordinatsystem, det betyder underlag, att få dem rätt, vrida in dem och så vidare.

Sedan behöver du ha någonting i plan. Ofta får du någonting platt, eller så vet du att här ska det bli ett hus, här ska det bli en parkering, en väg eller vad som helst. Sedan behöver du sätta höjden, och därefter behöver du göra det till en så kallad modell, en triangulerad 3D-yta. Efter det behöver du kunna ta ut volymer och jämföra med befintlig mark. Sedan behöver du kunna bereda det på olika sätt, det som är lämpligt, antingen på ritning eller digitalt, och därefter behöver du få ut det i verkligheten. Befintligt, plan, höjd, modell, volym, beredning och export. Det är någonting man kontinuerligt gör som anläggningsmätare eller anläggningsmodellerare. Tycker du att det låter intressant så gå in och läs mer. Den första kursen är lite av ett testmaterial.

Du kommer att få support utöver det vanliga. Du får också tillgång till ett fråge- och svarsrum, så att du kan ställa frågor om dina egna projekt eller när du kör fast framöver. Det erbjuds under ett år framåt om du bokar eller går någon av mina kurser. Du kommer att få en office time där du kan ställa frågor om vad som helst, om dina egna projekt. Och om jag kan svara och hjälpa dig med det så hjälper jag dig. Så, landskaparen.se. Men nu till intervjun, och njut av Gavins fantastiska erfarenheter.

Men nu till intervjun, och njut av Gavins fantastiska erfarenheter. Maybe you can just present yourself, your background, that would be nice introduction to the topic. Your background in GNSS and even before GNSS.

Ok, I am a licensed surveyer I been serving since high school, summer work. I lived in Australia at the time, West Australia, and it's a very active state for mining and other resources, so a lot of need for surveyors. So in high school, in summer work, in the far north, conventional surveying and celestial surveying to support these mining operations. And one year, we were recruited from high school to use the early, the early satellite navigation systems was known the United States Navy's transit system, which was purely Doppler. So it would take all day or several days to get a very precise position, which wasn't that precise, maybe 30 meters at best. But for those purposes, it was very useful.

So the previous summer, we did radio triangulation for these gas drilling rigs off the coast. And they would drop you on an island or some promontory of the land with a radio. And we did exactly that, triangulated the signals to the gas rig. But the next year was this Doppler system. Well, a strange looking device, very heavy. But we had to figure out where the satellites would be at certain times of day. So you get up in the middle of the night, and you record the Doppler and hand reduce it and come up with a position. But I had been reading, being a geek, I had been reading about this GPS system that was coming. And for military purposes, that was launched in 1978 with just a few satellites. By the early 1980s, there were 11 GPS satellites, and then it was commercially available.

Some of the early commercial receivers was like the Hewlett Packard HP4100, which I actually have my original one sitting in the office. Somebody sent it to me from the university. A very large device. How shall I describe it? Like the size of a dishwasher.

Wow.

It would use a lot of power and throw a lot of heat, so we could put the unit in the tent to keep us warm at night or cook our lunch on it. But the commercial ones like the Trimble 4000 series, still size of a microwave, 55 pounds, 25 kilograms. I've got to forget, I'm talking to Europe here.

Yeah, exactly.

I actually have one of my original ones sitting here. But things got miniaturized, and then when multi-constellation came along, that of course helped. So the early uses for the GPS were geodetic mostly that we would work with. Geodetic positions, either to do a pair of them to get an azimuth, which in those days, that could take days, because you had to figure out what days the satellites would be in view, and get your minimum of four satellites, and come up with a position. Well, at that time, for an azimuth, we could still use celestial observations, picking polar stars at the sun or the moon, and come up with an azimuth almost as fast. But the potential was there, and the equipment was very expensive. When Trimble sold their first systems with the 4000, it was a quarter of a million dollars, US dollars, and you got a pair of receivers with the antennas.

But we all know how the costs have gone down. Mainly used it for geodetic purposes until the mid-1990s when the real-time kinematic RTK was developed. Mainly the first RTK systems were German made, Spectra, under the banner of Spectra, which is then bought by Trimble. But that opened it up to much more surveying, land surveying uses. And in the end of the 1990s, we started to see white papers on network RTK. This meant you didn't have the limitation of baseline length. And I know your listeners are familiar with how this works, but I'll recap it a little bit.

Yeah, do that.

Real-time kinematic, well, you basically converge the solutions. The pseudo-range ambiguities of how many wavelengths does it take to get to the face center of the antenna from the satellite, but you do this double differencing for more than one satellite and also a base, and you put the base on a known point. The limitation is degradation over distance. In normal space weather conditions, about 10 kilometers, you could fix the ambiguities, meaning the wavelength is about 19 centimeters, so half a wavelength, you're within 10 centimeters when it fixes, and beyond that point, it uses the phase to come up with the more precise location, how far along that wavelength for multiple satellites. This is very handy, but to do RTK over wide areas or to make it more useful, when you put up a base, you have to set it on a known point, so you have to have the geodesy figured out really well.

Then the base, someone has to sit there with it, because it could get stolen. So the network RTK, it was reading about Terasat, which is based out of Munich a little bit. It's now part of Trimble. I forget the doctor that commercialized that. But you could put the bases 50 kilometers apart, 70 kilometers apart, something like that. And it was not quite as susceptible to the space weather in terms of that degradation over distance. A very common, you know, three or four approaches to network RTK. A very common one is they call the non-physical base or virtual or VRS, you know. And then there's master auxiliary. They all had their pluses and minuses, but the end result in precision, relative precision from the nearest physical base and then absolute accuracy, they would all achieve pretty much the same thing.

You're getting down to a few centimeters in real time. Then it became very viable. I was working for and still do work for in my day job, a very large utility covering a big metropolitan area of Washington state, the Seattle area. And to make RTK viable for wide use by a lot of peruse was putting up some permanent bases and running this type VRS type solution. Well, once we had got the first four initial bases up in the metropolitan area with a couple of partners, all the other jurisdictions in the state started to say, hey, can we connect? Can we add a station? Well, so then we formed a public-private cooperative, self-supporting and expanded across the state.

By 2009, we had covered the whole state, which is, the state is about a third of the size of Sweden. So we have 150 stations. We've been running since 2000, well, we first piloting in 2001. So we have 150 stations, you know, Sweden, in contrast, you're three times the size and you've got 470 stations. So it's a very tight solution. Yours is operated by Lent Materiet, if I pronounced that correctly.

That was enough.

In the early days of the networks, there were, you know, very well established networks in Germany. Each of the states had a network, but they combined together and it's called SEPOS. And then Sweden was getting theirs going, Finland, Denmark was completely covered. In the United States, a university in Minnesota started, there's one month before us, they had four stations. But we were an early RTN, and that's expanded worldwide. There's about, I used to keep track of every network, but there are so many of different sizes. You'll have sometimes a network for a mining operation or a region, a city, a state, or a whole country. Like Japan has 1,100 stations, I believe, put up for your geophysical sciences, but also network.

So that's where it's gone. We've been operating the same uses that Swepos, well, of course, land property type surveys, construction, precision agriculture, machine control construction by extension, science, science is big partners in our network. We have a lot of plate tectonic velocity and earthquake hazards. So they use all of our data to calculate the velocities, and some of the stations are in the state earthquake warning system co-located with seismic equipment. And now we have all of these autonomous vehicle type systems. There's so many competing ones now that are going to compete for the autonomy.

The SAP Corda network, which is Ublox now, the Trimble RTX, the various, there's three kinds under Hexagon, Veripose, Novatel, and then SmartLink, I think they call it. So we have a lot of these autonomous vehicle development companies that will partner in on our network or lease data or use it directly as a subscriber or partner. So that's kind of where I'm at. I still do my serving in my day job. And about half of my time is to operate this network, which fortunately with the software, a lot of it is automated.

Just to give some color, what's involved in maintaining or operating a network like that. Like I have no clue. I'm just taking everything for granted, you know. So this is like opening the curtains for how it works.

Well, you should visit Lantmäteriet in Gävle.

Yeah, Avle, yeah.

Okay. That's an amazing operation. Out of the networks worldwide, all those network operators kind of know each other and at media conferences. That network is recognized as one of the most sophisticated. They're doing a lot more with it, and they have full-time research people researching things like, well, interference, potential, geodetic research to link with other applications. But operating the network, I found that most networks are very similar in this, that there are like three operational levels. There's user ops, there's network ops, and then there's field ops. So going backwards from the field ops is just maintaining the sensors, building new ones, maintaining, upgrading stations. So I have to drive all over the state to like upgrade receivers or communications.

Usually it's cellular broadband modems to get the data to the processing center. So our processing center is in Seattle, and we run redundant systems. In Sweden, it all goes to Gävle, to Lantmäteriet. So you take in the raw observations from the geodetic grade base stations and you process that into network RTK, or the users can choose an individual station and do traditional RTK from that station. And you have to provide several different flavors of corrections. They all work very much the same, but because different ages of equipment, field user might have a very old rover that can only do CMR or older RTCM. And then you put out the modern corrections, RTCM 3.2 or 3.4 MSM.

It's called multi-signal message for all the extra constellations. So the field ops is maintaining the equipment and the communications and connectivity. The network operations is operating the network software. Now, there are several kinds of network software out there. Leica has Spider, Trimble has what they call Pivot, which I believe is what's being used in Sweden. Topcon, Topnet, CHC, out of China, there's network software. They're all really good software. And over the 20 years, I've seen the level of automation get to the point where I don't have to intervene very much. And the IT side of it, the servers, out of the network ops, it's more about babysitting those servers, maintaining their performance, uptime.

And then the user ops is interfacing with the users, the end users. This is, even though we're a not-for-profit network, we're self-supporting, so either someone's a partner, they have invested in the infrastructure or operate a station, and they get logins in exchange for doing that. That's most of our users. And the non-partners pay, well, I guess you could call it a subscription. And that revenue goes to funding our operation. So the user ops is maintaining accounts, logins, and then support. Now, we don't support their hardware and software. If they have questions about their particular brand of Survey Rover, that question needs to go to the dealer that sold it to them. But we answer the most common questions we get are somebody says, is my login okay?

Well, they can use an app on their phone with an Ntrip client and determine that. Is the network okay? Well, they can go to the web interface. And we get a lot of strange questions. Also, we get treated like a general surveying support line. Someone will call up and ask me how to configure their total station. And I go, that has nothing to do with GNSS, but or how to install a game on their phone. We get the questions about which dovetails into your topic. They go, oh, they're jamming over in Eastern Europe. I mean, is that jamming us? And I'm like, no, that's kind of impossible. But we get those kind of questions and we have ways to show them that they can check online all kinds of ways to prove that, no, that's not really happening.

Yes, that was a nice overview and that makes us a nice transition to the jamming topping if we start with that. Like, how worried should one be for the vulnerabilities of GNSS and what are realistic threats? Have we learned something from the wars that have been and like from the military side or jamming side of things?

Well, you know, jamming is real and it happens every day typically on the battleground. There's all kinds of ways that interference can happen. Although, they are often overstated, but the potential is there. There's inadvertent or accidental interference, and that we can talk about, you know, what the typical situations are. Very rare, though. The battlefield jamming is very specific, targeted. It's limited in area for lots of reasons. And in the battlefield, there are lots of ways to mitigate that. The military have ways with specific antennas that can do directional blocking of signals with filters. The soldiers know that they can actually stand with their, wrap their stomach around the antenna and block in a direction.

There's all that. And that does happen, but not as much as people think. There's a great website that's called gpsjam. org, where you can see the hotspots around the world, where there's jamming going on a regular basis. And that is very limited. And going in even further, it may only be a few miles, where they have a tactical reason why they might jam. So, it's not that widespread. Before I go into that topical jamming, we'll talk about the inadvertent interference. You know, it's radio signals, very weak radio signals. And there are certain things that can cause interference. At the end of a major runway, where they have altimeter radar, you know, I know that at the end of our big airport here, if I park down at the end of the runway, when a plane comes in and they turn on the altimeter radar, well, that's not only interfering with the GPS signals, but my cellular signal and other, any other radio.

And it's very brief when the plane goes over. There can be some industrial sites where they've got persistent RF, you know, in frequencies that can interfere. We've only identified a couple in our state, very localized. Yeah, there's accidental. You know, there was a case in San Diego where the Navy turned on accidentally something on a ship that was in the harbor, and that jammed a lot of GPS. Something they would only really turn on when they're out in the middle of the ocean. Another case where a small television station, and they wired the antenna bad, in a bad splice, it sort of sprays RF out broadly, and that was interfering. That's the inadvertent. Also extremely rare. You know, understanding the economic impact of need for satellite communications and positioning.

This is tracked. There's a reporting center in the United States run by the US. Coast Guard. And I visited them and in regular contact with them, and they really don't get very many reports that turn into anything real. People will think they are getting jammed, and it turns out that they are not. And there's some deliberate jamming, what they call the trucker jammers. So, you know, somebody working for a large trucking company does not want their boss to know that they made stop somewhere, and they can buy this little jammer, they plug into a cigarette lighter, buy it online, and it can cause jamming in a very small area, mainly to jam their own GPS in their own trucks, so they can't be tracked. Well, there have only been a handful of cases where people did this, and one highly publicized case was a trucker jammer going by an airport in New Jersey which interfered as it went by.

Well, they got busted for that and that was like a $36,000 fine. So people aren't really doing this. Deliberate jamming is what's their end goal? Why would they want to do it? And there's not that many instances of why somebody wouldn't want to deliberately jam, considering the consequences. So that's also very, very rare. When it comes to the battlefield, as I said, it's very localized. Eastern Europe, Ukraine is highly publicized now. Around the Donetsk region, that'll show up that there's jamming on fairly consistent, very small areas, though. And the areas are so small that I think I sent you a link to something. So there's a permanent geodetic station. There's permanent geodetic stations all over the world and several in the Ukraine.

And I sent you a link to one that's very close to the Donetsk region, where I can go online and the scientific community stores the data from these for plate tectonics and atmospheric studies. And I can go on any day and get data from it, analyze the data and it's very clean. I think you could see from the graphs and the link I sent you.

Yeah, I had a look and I'll make sure I'll attach it to the episode link as well.

Yeah, and what it shows is that, yeah, right next door is all of this military activity going on, but it's not affecting that station. So the perception that people get that, you know, the jamming that is going on in Eastern Europe and the Middle East, you know, around Syria, that affects them in Sweden or the United States is no, that's just not possible. There was all the fear about that Russia or whoever might want to destroy the GPS satellites. That would be an act of war that would have bigger repercussions than just what's happening with us in the field. And also, there's multiple constellations. Somebody would have to destroy all of them to cause this unavailability of GNSS completely. Well, and that's different international players.

Do they want to pick a fight with Europe and Galileo? Do they want to pick a fight with China and Baidu or IRSS and India? That's kind of highly unlikely that someone would want to do it. Jamming them broadly would, well, there's, I got to send you a link to another analysis that, why Russia doesn't jam GPS as much as people think is, well, they're using GPS. They've found downed aircraft and military vehicles that have commercial GPS devices taped to their dashboard. You know, they rely on...

They need it as well, basically.

Yeah, they need it as well. You know, soldiers, you know, many militaries around the world are buying commercial units because they're really good and cheap and available. And the military supply line and quality is not always up to scratch. So, yeah, no one wants to shoot their own foot. That's why that specter of doomsday scenario of destroying GNSS is highly unlikely to the point of kind of ridiculous. But these are vulnerabilities, and our reliance on these is very, very high. Well, that's also the reason why there's the technical factor, but the human factor. And the human factor is human activity realize on it so much, nobody wants to screw up their own usage.

So that makes sense. That avenue is probably quite unlikely. If that would happen, it would be like a major global act of war or like nuclear escalation or something in that. More like in a grand scale of things, like it would be a conflict of biggest scale, world scale, and then there's other problems in that case.

Yeah, it's like solar activity, which does affect if it was a big enough solar event that it crippled GNSS, well, that would be the least of our worries. It would be crippling so many other aspects of our wired society, communications, electrical grids.

Yeah, so let's talk about that a little bit, because I read a report of, was it the UK that had a report that about the impact of GNSS going down and possible causes, and the solar storm were identified as one possible cause that GNSS could be going down for a few days or five days or something like that. What's your thinking of that? Or like, what's the background? Like, how realistic are solar storms and their impact? Like, let's say, I've been reading about this Carrington event in 1859, that was the strongest solar storm that we have recorded in the modern day. And then there was the Quebec storm and took down the electrical net. Like, what would happen with the GNSS systems today if something like that would happen?

Yeah, no, that's very real. The solar activities, you know, the coronal mass ejections and, you know, the solar flares. In the time we've been running our network, there have only been five days in 20 years where somebody might not want to use GNSS. It was just the quality got bad. It didn't knock it out completely. Those events that you talked about, it would not only be GNSS, it would be other things, our cellular communications, electrical grids. That's bigger worries. But we haven't had one of those in modern times that affected very broad areas. That's just a reality that there's always that specter that could happen. What would we do if we were down for a few days? Now, the solar max, the solar cycles, we've been through a couple of those already.

And during the solar cycles, well, the GNSS satellites themselves are very hardened and they have mitigation measures. They're going to keep working, but the signal on the ground, that could be affected. Most networks, we monitor the ionospheric activity, whether it's heightened to the point, and the user can look at a simple graph and go, okay, today maybe I want to use my total station. Like I said, that's only been five days. The mitigating for that in a network solution, it can model out a lot of that very well. If somebody is doing single base, that degradation over baseline length from the base would just get shorter and shorter. You know, the 10 kilometer nice day range is very good. No one can really predict when these would happen very well.

I see warnings every day from different space centers around the world saying, oh, coronal mass ejection, solar storm. For the scientific community, that they're much more sensitive to what that might mean for their work with their other satellites, weather satellites, other communications. Very sensitive sensors for different purposes. But I see those warnings and I look at our graphs and I go, it's heightened, but it's not in the danger range. Typically, I never hear anything from the end users or our own monitoring equipment. I can look at data from other networks or the scientific. You know, there's like 3,000 or 4,000 permanent stations in the United States that we can look at data. And so far, maybe we've been lucky, but a solar storm warning comes in, we don't really see any effect.

Maybe it takes someone a little longer to fix their solution. Maybe their quality drops a little bit. But I don't know, maybe I'm too optimistic. But for the most part, unless there's one of those massive events that's gonna affect everybody and everything, no, it's not a huge worry, like for the surveyors in the field.

If there would be a bigger case, like what are like suitable mitigations? Because it's a low risk, let's say it's a low risk event, but the high impact or low probability event, but more high impact. Like, how would you start thinking about, like if you're running a bigger company and you're responsible for that work goes on? Like, do you like would estimate it as such a small probability? Or like, the day it happens, it's everybody, it's like a natural, like a big storm, like everything in the society goes like into, nobody expects normal service if something like that would happen. Or is it something that you need to prepare? Like, I'm thinking like, should you put controls on construction site, regardless, should you have like total stations in like excess capacity or like extra, so you can actually use total stations if you're usually just using GNSS?

Like, what's responsible level of preparedness?

Putting local control is an essential practice anyhow, even if you're using GNSS, you want to be able to check in to calibrate things like that. So, typical operations there, I mean, there's some people that just use the GNSS for everything and rely on it, which is not a good practice for other purposes, their own quality control. Construction sites will have local control. Then, surveyors doing property surveying. They're kind of using it in a relative manner anyhow. It's the relative positions of those boundary markers or boundaries from each other that's more important than the absolute position. So, you know, okay, big solar storm can't use the GNSS today. We're in a purely relative mode with the total station. And then later, you could go back and put your geodetic positions on some of that control.

That's it. You just got to be prepared for the worst. People are going to lose more days of work for weather events, storms, than they are space weather. And because GNSS can't do everything, there are so many activities that have terrestrial alternatives in their toolbox, toolkit. They have the total station. They have the scanners. Falling back to that, yeah, there's going to be people running around throwing their hands in the air, going, oh my god, what the hell do I do? But most are just going to go, oh, okay. All right, well, let's pull out the total station and shoot that and then this will pass. Definitely, you do want to put local project control, no matter what you're doing. Yeah, it's to calibrate to that locally, where those relative distances and positions in whatever you're doing is more important than anything.

Absolutely, I agree. I'm just thinking since GNSS and RTN is so easy to use and so cheap to use, you have a new generation of users that are more treating it as a black box and it just works. Maybe they don't have the knowledge or actually understand how it works or how to do it. Otherwise, you just have like the magic wand and you walk around with it.

Yeah, that is the, well, you know, surveyors here, especially the older ones, my age or older, you know, always we talk about, oh, they're just button pushers, you know. Well, when that button means so much efficiency, it's like, it's a good practice to use it. Because for the most part, except extremely rare occasions, it is going to work and it is going to be precise, you know, and efficient. And that's it. It's the efficiency, the risk factor. Can I afford to work with legacy methods? Can't stay competitive, can't be efficient enough. With the offhand chance that there may be times when you can't use it. Well, if I have a snowstorm, I'm not going to be able to go do it as well. Not much you can do about that. There is a lot of education needs to go on with newer users.

I am very happy to find out that the level of sophistication of the users is pretty high now. They do get it. They do understand the limitation. They do understand a lot more about how it works than, ironically, than the older surveyors first started using it. You get a new crew member who's never surveyed before, and you put them in the field with the GNSS unit. They have an idea of how GNSS works because it's part of their daily life and everything else they do. The map on their phone, the location-based applications, understanding that, oh, okay, it gets really bad around a bunch of tall buildings, multi-path. They don't understand multi-path, but they go, yeah, no, no, it's signals bouncing off stuff. I'm under a big tree. That's not going to work all that well. It's not going to work indoors. They get it, and the users do get it a lot more than in the early days when it was completely puzzling to everybody.

In a way, sometimes a survey or my age that's used a lot of legacy stuff, we have a lot of baggage to carry along. We keep picturing things in the same way that we would do terrestrial, and it is a little different. We need to do a lot more education for sure.

Well, that was a really nice point that, yeah, everybody uses their smartphone, and you have a sense that, okay, I might be on the wrong side of the road, or here it doesn't work, and that's true. So you have like a, most people have a sense of how it works, because it is a part of the daily life, even if it's not that accurate, a survey grade, but the same principle applies. Is there anything else about GNSS vulnerabilities that you've been thinking about, or like, what's possible mitigations, or any thoughts you have about GNSS vulnerabilities?

Well, you know, the drive for alternatives is a high priority in different countries. You know, the agencies that represent large concerns, departments of transportation, they're doing a lot of research on alternatives, or mitigation, or reliability measures. Especially, you know, with the autonomous vehicle stuff coming, it's not going to come as fast as some people would like. You know, a total free-for-all on the highways where you self-driving, that's going to take a while. There's a lot that needs to go on, the vehicle-to-vehicle stuff, driver assistance, but then vehicle V-to-XV to everything else. You know, like beacons, or alert systems about, oh, okay, yeah, my car can navigate a lane very well, using many other sensors besides GNSS, but what if it's a construction site and they move that lane temporarily?

You have to notify people. But alternatives for navigation, you know, there's a UK, a US and other countries want to boost up other things like the eLauran. You know, Lauran has been around a long time for navigation, but the eLauran, they want to do this more modernized version. Well, that's signal-based. That would also be susceptible to a solar storm. That's great. Ground-based positioning systems, you know, beacons, a lot of work going on in doing those where you have transmitters on towers around a city with atomic clocks in them, and you do ranging that way. Well, that's signal-based as well. Pretty much any navigation and positioning system that can work in a wide area for mass applications is some kind of signal. A lot of work going into low Earth orbit satellites for positioning and navigation at the LEO.

Well, there's advantages. The LEO signals are stronger. You know, they're lower orbit. You can get geometry advantages. You get signal strength advantages, harder to jam, susceptible to solar storms, absolutely. And LEO satellites, well, the other thing about those is they're low orbit and they don't stay up there that long. You have to replace them frequently. You know, people have these visions of they go, oh, there's all these low Earth orbit communications satellites like the Starlink or something like that. And it's like, no, even those providers aren't really promoting positioning from it because communications satellite, well, and they're very small and have a power budget issue. And to do positioning with them, you need to dedicate hardware on them. So you can't just go, oh, my God, there's 700 of these communications satellites.

Let's do positioning from them. Well, OK, you could do Doppler. You can do primitive ranging. It's not very precise. There are outfits like Zona Space that are going to put up LEO dedicated for positioning. But that's the package in it is specifically for positioning. So this provides you some alternatives, but then again, you're back to signal-based. Really, how shall I say, blue sky mad scientist type thinking that the quantum sensing could provide advantages. Quantum navigation device, which is inertial using subatomic particle behaviors. There's some promise there. It won't be super precise, but that would be independent of any of the signal ranging. Then again, high electromagnetic disturbance could hurt that.

A couple of days ago, I was at JPL NASA, so I was interviewing some of the people there about the quantum sensing for gravity meters, for clocks and for the navigation. And it was really fascinating, but they're a long way from something small enough, smaller than a backpack. But there's potential there. That would be independent of the signal in space type situations, but we're quite a ways from it. Yeah, laser-based. Then you're kind of limited to EDM line of sight. So you're back to your total station. So there's no magic bullet that I can find, and I look all the time on the horizon. GNSS is there. It does have vulnerabilities, not as likely as some people would like to think. And then, of course, there's money to be made scaring people about it, trying to sell them all kinds of alternatives.

But because it is so widely used globally by so many people, and the economics that relies on it, that actually becomes a strength against a lot of the potential risks.

It is a fantastic system, and when you read about it, how much time and thought and resources go about to keep everything maintained, and it's a robust system, surprisingly so.

Surprisingly, yeah. Yeah, it is. It didn't used to be. It got so much more reliable now.

If we leave the GNS's vulnerabilities, I was just, if we can talk a few minutes about the quality of the new, let's call them low-cost receivers. I know you've written a few some articles about it and tested, but you know, it's here if you go to any trade show, you have seven or eight manufacturers or resellers selling GNS's rovers. And sometimes it feels like the only price difference is price or actually, or rather you don't know what's the difference. You can see the price difference and there's color differences. But what is there to say about the quality differences and accuracy differences? And like how should a common survey or a common man like evaluate different options? Like you have the like and trim well, most expensive variants to MLEADS and very low cost, like 10, 15 times lower cost receivers.

And if you read the specs, it's almost the same.

Yeah, it's great that there's so much more variety now and so many more options. And it is possible to build a very inexpensive GNSS receiver now. To some degree, the signal is just the signal. And what you do with it, you can limit costs by very small, inexpensive chips. The push for consumer and prosumer level uses, particularly like on drones, people make very small chipset. And for the autonomous systems and Internet of Things type applications, robotics, location-based services, the chip can be very inexpensive. And there, I got to be careful not to talk about brand because I write about stuff and I got to not make any judgment.

Okay, you've got different levels of the GNSS receivers, and the antennas, which that makes a huge difference. For an application like I'm going to have a drone, and the chip that goes on the drone, the GNSS that goes on the drone, it's clear sky up there. And it doesn't have to be all that great. The antenna doesn't have to be all that great. You know, I'm not going to have multipath up there in the sky. Well, flying it around a bridge, okay, different subject altogether. But the chipset and it can be multi-constellation, but they might only pick a couple of signals per constellation. You know, they might pick an E1 and an E5 from Galileo and L1 and L2 or L1 and L5 from GPS and so on. So you only have a few channels needed. It doesn't need all that much horsepower to process that.

So start to go to the other end of the scale, the high-end boards. And for a long time, there were a handful of manufacturers and they would produce boards and sell them as OEM, other equipment manufacturer. And then you had a whole bunch of other rovers that of all these different brands and colors, you know. And for a while, there are three quarters of all rovers in the world for applications like surveying. The boards came from two manufacturers and then three. Well, we stepped into an era what I call the fourth wave of GNSS, and that was coincident with the multi-constellation. If you are going to use four or five constellations, and they may have not only dual frequency, but three, four or five signals per constellation, and a lot more satellites, you're pulling in 50 satellites with three or four signals from each.

Your board, your processor has to be able to handle that, putting everything under one giant filter, and combining those to come up with the best quality. Only a few years ago, some of the manufacturers, big name ones, had rovers that would track all those constellations, but couldn't really use them all together in one solution. Their next model, they would double their processing power. So then you could, and that's the high-end stuff. This is the really expensive survey rover. And in some aspects, it's all well worth the cost, because you can get the reliability, repeatability and usability in mixed environments that the very low-cost ones just can't. I can take, and I have taken, some really inexpensive rovers. And I'm out in the open, and everything's fine, and I can get my centimeter.

Well, I can take a 25-year-old GPS-only rover, buy off of the Internet used for a few hundred dollars, and I can get the centimeter with just the GPS signals out in the open. That's the difference. When you got to mitigate for multipath, the less expensive ones often are not going to work all that well. They're not truly going to do full constellation for like working in tree canopy very well. It goes down to a quality thing, and if somebody's, you know, high stakes serving the least expensive, there's a risk factor there that in that particular environment, that might not work all that well. So it is great to have all that variety, though. And the monopoly on high precision that a handful of big companies had, well, that's gone.

There are third-party manufacturers in different parts of the world that are actually building their own boards. They don't buy from the big ones anymore. They build their own boards, which performance-wise can match them spot on.

Hmm. That's really fascinating development. How well, like, you can see ads about this works really well on the canopy, best on the canopy and so on. How good can you expect it to be with a canopy cover? Because I can see that on my rover. If I go with a new rover under a canopy, you can kind of get a good, or the coordinate quality values are quite good. But like when I compare the data, it fluctuates more like when you and the coordinate quality number would imply.

The best in canopy is kind of humorous. They're always claiming, we're the best in canopy. Well, the canopy is the canopy. No matter what you've got, you are going to have limitations. You're going to have less consistent. So from our end users and from, I did a lot of independent testing for different outfits. And out of my own curiosity is there's no magic bullet. Nobody has the magic bullet. In general, the higher quality of the GNSS unit, receiver, antenna and how they process it will get you better performance. The high-end units now, as I said, the fourth wave, doesn't matter whose it is. The ones in the past few years that can encompass all of the signals are performing under canopy.

You've noticed it so much better than even a few years ago. But still, you got to be diligent. You're surveying under canopy. You got to... everything is suspect. Depends on the canopy too. We got these Washington state, the western half of our state is like parts of Sweden. We got these great big evergreen trees, deep dark, ork-filled forests, right? Almost nothing's going to work in that. You can mitigate by getting the antenna up higher. It's really fascinating study that Japan did to universities, did a study for their forestry industry. And I talked to the folks that did it. And they said, you know, it was really funny that just raising that rod, another couple of meters made all the world a difference in what they could do in the thick canopy. The newer RTK engines that put everything under one giant filter and can pick the best of those signals from all of the ones getting through the trees, improvements.

For instance, the signals themselves. In Galileo, you have the E5A, E5B that can be processed, the AltBach, they call it. That is a great multipath mitigator, which helps you in the trees. So the new gear with all the signals and a hell of a big processor we're seeing usability in a lot of canopy, but not all of it. And then your base has to support those signals as well. People buy a new rover and use their old rover as a base, but it doesn't support all the signals. That's a common one I get with the end users. Another angle to this, no pun intended, the inertial units, like in the tilt compensation, no calibration tilt compensation. One thing about that is you're using trajectory to come up with an orientation trajectory and the position to come up with the orientation for the tilt.

The magnetic really sucked. That was tilt that was kind of a joke for years. Well, the other thing that trajectory does is like with the mobile mapping systems or the airborne systems. That actually helps you identify outliers. So if you have inertial units, and they haven't all implemented this in the RTK solution for this purpose yet, but a lot of them are, and I got to try out a couple, that actual movement helps you identify outliers in what you're doing. So under Canopy, if you're moving it a little bit, it helps you identify the bad positions. On taking that a step further, cameras or SLAM LiDAR, there are going to be rovers with very small, solid state LiDAR units on them, or the types of integrations like on the iPhone, where you're also able to use that LiDAR positioning for stabilization of the position.

It's going to be able to identify the outliers. There's some exciting things going to be happening in rovers in the near future, where we'll get a little more gain in the canopy or urban canyons through these additional sensors. Multi-sensor integration is where it's going to go.

Fascinating. I can hear that you have much to say.

Sorry.

No, no. I'd love to.

Don't get me started. Yeah.

Exactly. It takes an hour to get started. Then it's good. Yeah. But you have a workday ahead and I have a screaming child downstairs. I can hear here.

Oh, yeah. Yeah. Yeah. I shouldn't keep you long.

No, no. But I really appreciate this and I could listen for hours. Perhaps let's see what we can do sometime in the future. But let's wrap it up by, because you're quite a prolific writer I found on the Internet, or you write stuff. Where can we find stuff to read about you that you've written, like magazines or places? So where can we follow you?

GPS World. You know, LinkedIn, I usually will link my articles in LinkedIn, no matter who they're for. It's funny, I got started in writing. I didn't start it until about 25 years ago, but I started in writing because I had a project that had questionable funding, you know? And I thought, you know, a big glossy article about it. And then the funders go, oh, they feel part of it, you know? So there's gschrock. com. I put my writing samples, but LinkedIn, find me on LinkedIn. And then I link mine and other people's articles that are relevant to surveyors in our technology.

Awesome. I will link to those. Thank you so much for your time. And I really appreciate all the insights you shared.

Okay. All right. Take care.

Take care.

Thanks.