Evolution of Technology | EP 21

Carson: Welcome back to our Link Labs podcast, Get Linked. I'm your host Carson Garner. We've got another great episode planned for you today. Today is actually our 21st episode. And on our 21st episode, we have none other than Chief Engineer Mark Bloechl coming back for an episode on the evolution of location technology. As you all know, technology is ever-evolving and it's always rapidly changing and advancements are being made so quickly. Today, Mark and I will be discussing some of those advancements and how they are impacting the world on a global scale.

Carson: Mark, thank you so much for joining me for another episode of the Get Linked podcast.

Mark: Thanks. Thanks for having me.

Carson: So you've been with the company for a while and I know we've actually had a prior podcast. I think it was episode nine that we had you join us. And on that podcast, we had you introduce yourself a little bit, but we might have a few new listeners this time. So could you reintroduce yourself and tell us a little bit about your experience at Link Labs?

Mark: Sure. Yeah, so I've been with Link Labs since the beginning: a little over nine years now, back when we were focused initially on building out an IoT network. And then, kind of real quickly, we recognized the opportunities in RTLS systems and positioning systems, and so we kind of pivoted and focused on that. And so ever since, I've largely been working on our edge location technology.

Carson: Yeah, yeah, and you've been doing some amazing things with UWB and the extremely high level accuracy tracking technology that’s just amazing to see. I guess not every location technology works the same though, right? So what are the most popular techniques that people use for this?

Mark: Sure, so I kind of group them together in three broad categories. You have basic signal strength based positioning. So these are things like BLE proximity, which we'll probably talk about more here, or trying to do triangulation from signal strength. Then you have your time-based systems. The one everyone is most familiar with is probably GPS, that’s where you're measuring the time it takes a signal to either go to some known point and back or just for it to transit from that point. And from that, you can figure out how far away you are from that point, and then you can position yourself. Ultra-wideband is also a good example of a time-based system. And then finally, the third broad category is angle-based systems. Some people may be familiar with this, with the latest version of Bluetooth, that has an angle of arrival feature. Basically, with this technique, you have some known location and a signal comes in. You have a receiver at that point, and it's able to figure out the angle that the signal arrived at. And if you get enough of those observations, you can pinpoint where the signal propagated from. So those are kind of the three broad categories. There's obviously more niche ones, but I think this probably covers most of your positioning systems in use.

Carson: Yeah, yeah. So, and I said that you've been with the company for a while, but that actually means that you… I mean, you could say that you pretty much helped create the company, right? You helped with the technology. You studied at John Hopkins. Is that right?

Mark: Oh, yeah, I got my masters at Johns Hopkins. Yep.

Carson: Great. Great. Yeah. So you've been around location technologies, you know how to develop them, how they work. Let's start with Bluetooth Low Energy. And that's a proximity-based location technology. Can you explain how it works and how the technology determines location?

Mark: Sure, yeah. I mean, probably the easiest way to explain it is imagine you're in a crowded room with conversations going on all around you. You could kind of infer that if you're hearing a conversation louder, one conversation louder than another one, then you're closer to the person that's speaking.

Carson: Yeah.

Mark: And that's exactly how it works. The tag… Or, you know, at least for our system, our tags, you know, scan. We have beacons at known locations, and those beacons are broadcasting. In BLE, it’s called advertising. But they’re basically just broadcasting a signal saying who they are. And then tags scan, you know… The simplified version is the signal that's the loudest is the one that they assume that they're closest to. It's a little more nuanced than that, but that's the gist of it. And that is somewhat unique even in proximity-based systems. A lot of other systems actually do the opposite or flip that link around, so the tag is transmitting and you have beacons that are listening. And then they radio those observations back. And we did this because it allows us to actually get better accuracy at a lower cost. In a proximity-based system, the beacon is the location, and so, if you wanted two desks side by side to each be their own location, you need two beacons. It's a much lower cost to be able to put down just some battery-powered beacons for our system versus receivers that actually have to listen and then flow that data back and often be powered. So that was kind of the tweak we put on proximity bait.

Carson: I guess how do you flip that, from like a technical standpoint?

Mark: Yeah, so it's a bit of an atypical way of running the system. The tag has to periodically open a receive window. And there's a little more complexity code-wise to make that work. We took our algorithm that we used to… At first we tried running a system where the tag transmitted and beacons received, and we had our location algorithm up in the cloud. We basically ported that down into the tag, so we rewrote it to run on the tag efficiently. And yeah, so it just opens a little receive window, catches all these observations, and then turns them through this algorithm, and it spits out which one it thinks it's closest to. Which is actually another reason why the system is more accurate. 

Carson: Yeah.

Mark: You know, it's listening and receiving all of these observations at once at the same time, effectively at the same time. In other systems where you have that flip link and all that data flows to the cloud or some central spot, you know, there's no guarantee that these different receivers are going to receive the signal at the same time.

Carson: Yeah.

Mark: So that was actually why we did it that way, just trying to time synchronize everything. So, yeah: much simpler, lower cost, higher accuracy.

Carson: Yeah. And that's what we want, right? That's the best thing to sell to the customers, because that's what they need. A lot of customers, especially in manufacturing and other indoor environments are looking for high accuracy and affordable cost. And so any efforts that we can do to make it that way are going to be revolutionary to this space. But let's move forward a little bit and talk about phase-ranging technology. How does that work and how does it differ from that proximity-based technology?

Mark: Sure, yeah. It's very different. As far as I know, we are the only providers, only people using this technology. We've kind of pioneered it, really.

Carson: Yeah.

Mark: And actually, it's not a proximity or signal strength based system at all. It's actually doing… It's a time-based one. It's doing ranging, so it's measuring the distance between a tag and these beacons. Traditionally, the way you would do this is, for example, with UWB, you transmit a really wide bandwidth signal. That can… You know, the issue there is that it can get power hungry and it requires specialized chipsets for UWB and even UWB radio. And so it can be a little prohibitive for low-cost tags. So what we did was, what we do instead is kind of a little more niche technique. Basically, instead of measuring the time it takes a signal to go to a known point – you know, the beacon and back — we actually measure the signal phase.

Mark: If you remember from trigonometry, it's basically you have two sine waves and one shifted from the other and that's the phase. I mean, that's exactly what this is. So the tag sends a tone, a sine wave you can think of it as to this beacon, which kind of bounces it back, and the tag measures that phase shift. And phase is just another way of measuring time, basically. But these little chipsets – Bluetooth chipsets – are really, really good at measuring phase accurately because that's how they communicate. So we're kind of leveraging that. And we use that to develop our proprietary phase-ranging technique, so it runs on standard off-shelf Bluetooth chipsets. And we can get on the order of two to three meters of accuracy off of that, better and better environments. But unlike the proximity systems – you know, when the tag calculates its location – we've kind of kept that flip link architecture so the tag calculates location instead of sending. You know, an idea of a beacon that is closest to it actually sends coordinates. So it's just X, Y, and Z positions and so, you know, we can display that dot on a map, you know, kind of anywhere. We're not constrained. 

Carson: So in the last episode, we discussed ultra-wideband and that's kind of what we wanted to hit on here because after all this episode is the evolution of our technology. And I think UWB or Ultra is our evolution, where we're at right now. So how would you say that UWB compares to the others that we've discussed?

Mark: Yeah, so, you know, like I said, UWB is a time-based system. It kind of takes that classical time-based approach where it transmits a really wide bandwidth signal. When you're trying to measure time, you know, the wider your signal bandwidth, the more accurate you can measure. So maybe it transmits a 500 megahertz wide signal. And so, you know, you can measure down into picoseconds, tenths of picoseconds. And so, you can get down to centimeter level accuracy with it. The tradeoffs are: it is a separate chip, it's a separate radio, separate antenna. And it's something we kept our eye on for a long time, but we didn't really feel like it was right to jump into until just recently, like last year, I think. The latest generation chipsets came out. They were much lower power, kind of easier to use. And so we effectively pulled them into our system so our systems can actually support mixed installations. So you can have phase-ranging tags working with the same beacons that can also support our UWB tags. So it really gives customers a lot of flexibility in terms of cost, accuracy trade-offs, and things like that.

Carson: Yeah, but UWB would be more cost-oriented.

Mark: Yes, it’s an additional radio, you know, additional chip. So, yeah, the cost is a little bit higher. Like I said, it has come down a fair amount, you know, to companies like Apple getting into UWB. I think that's kind of driven the cost down. So, yeah, the radios are definitely more cost effective than they used to be just, you know, five, ten years ago.

Carson: Yeah, but I think that's the beauty of having a company like Link Labs, because we have options. And, you know, if you don't need that type of granular detailed level tracking, you know, we can go with an XLE or a BLE. But if you do need it, we can, right? And that's having that ability.

Mark: Yeah, I think that was most exciting for me about pulling in UWB is we really have the full range.

Carson: Yeah, everything.

Mark: We have the best. And so yeah, there's really no application we couldn't conceivably support.

Carson: Yeah, and in sales, that's key, right? Having that ability to present different options, right? And I think, you know, some companies, they won't go with the Ultra infrastructure just because of how expensive it is. Or they don't need that level of accuracy. But there are ones out there that do, right? And having that option is key, like you said. So, each of these different technologies, they work a little bit differently, they serve different purposes. Could you give an example of when you would want to use one versus when you would want to use another?

Mark: Yeah, yeah, absolutely. So, you know, what we've found, given that we now have all three of these kinds of technologies, what we found is that, there is still a place for proximity positioning. It is low cost, low accuracy. The infrastructure is really lightweight. The beacons are all battery-powered. And so, where it really seems to excel is if you have an environment that has lots of rooms or discrete locations – things are either here or they’re here, that kind of thing – and you're able to put a beacon there, then it works. It works well. And if you don't need really precise positioning, you know, we always say the accuracy of our proximity system is basically plus or minus one location beacon. You know, if it doesn't pick the right one, then odds are, it's just off by one. So, you know, if you can tolerate that level of precision, then proximity can be a very good solution.

Carson: Yeah.

Mark: Which is why we still have it. For phase-ranging, that's the next step up. It's still low cost because it's effectively the same hardware. We don't need any extra chips for that. It uses the BLE radio for the phase-ranging. But you do get better accuracy, like I said, generally 2 to 3 meters in nice environments where you don't have a lot of metal reflectors or things like that. You can even get down to a meter and a half. And where that kind of shines, what we found is if you have kind of open environments where there isn't discrete locations where things are at – you know, you've got a big open warehouse or open work area or something like that – you can still do route tracking within rooms and things like that as well. But if you have that big open area, it's difficult to kind of use the proximity based positioning and place beacons all throughout. Especially warehouse environments. You have a battery powered little puck stringed somewhere, and they could get knocked off or whatever, you know. For our phase ranging technology, we generally mount those beacons on the ceiling, and so they're out of the way. They’re powered, and you never have to worry about batteries. Uh, you know, it's daisy chain, low voltage power. So that's kind of where it shines, I think. And then UWB would just be the next step up. And like I said the infrastructure is the same; you know, the beacons will be the same. They'll work with both radios, and be able to support both. And so it's not really much extra cost or any extra cost to support the network. It would just be the tags. So if you have some assets that you really need precise positioning on, you can put UWB tags on those and get sub meter accuracy. If your environment is really challenging, with lots of multiplat reflectors and interference and things like that, UWB will get you more accurate results. But then if you do have lower cost things that you just kind of want to know roughly where they're at,  you can put those phaser tags on those and work together.

Carson: So you can mix and match like in a single and like one floor plan. Interesting.

Mark: Yep.

Carson: Interesting. And you answered one of my questions that I've had on a personal level for a while, because I've wondered like… Whenever we're selling the technology to customers, it's less hardware, but it's a little more expensive. But then when you're on the BLE side, it's more hardware and it's less expensive. So that was the question that I had, like, you know, how does that work? Why is it like that? But that makes a lot of sense. So with UWB, would it be, you know, more hardware or would it be less hardware?

Mark: Well, actually, the idea is we're going to be putting UWB radios in each of the beacons anyway. And so it's kind of going to be baked in and we'll have, you know, ultimately we'll have one beacon that supports both.

Carson: Okay.

Mark: And there's several reasons for that. You know, one is that customers may not know exactly what they need at first, and maybe they're okay with the phase ranging, but at some point, you know, they might want to be able to do more precise positioning. And so, you know, having that one SKU that can support both is good. The other thing is, we're also researching and working on some kind of quality of life enhancements for the system. So for installations, if we have these precise positioning radios on all of the beacons, then for example, they can measure distances to themselves and do a certain amount of auto placement so you don't have to go around and measure all of the beacon locations.

Carson: Yeah, that's convenient.

Mark: Yeah, yep. So, yeah, that's kind of the idea is that, on the infrastructure side, it'll just be one, you know, one device.

Carson: Yeah, yeah.

Mark: And the tags are what differentiate the technologies.

Carson: Yeah. So far we've discussed all the indoor location technology, everything that you're an expert in. But let's briefly switch over to the outdoor location technology. Kind of more of my ballpark because sometimes I'm not as technical, but are there any technologies comparable to the ones that we've been talking about on an outdoor scale or that a customer could use outdoors?

Mark: Yeah. Yeah, I mean, the obvious one would be GPS. I think I mentioned that already. That's a time-based positioning system, been around a very long time now, and works really well outdoors. The other one that we utilize is Wi-Fi positioning. That's essentially tags doing scans, looking at different BSS IDs they see for Wi-Fi networks, which if you're in a crowded area, city, something like that, you might see a lot. And then they go to an online database and do a location lookup based on that. And you can get pretty decent accuracy with that. So we actually use both of those technologies in our SuperTag product. It also supports cellular ID as well. Yeah, so it has a plethora of options. You know, and the other neat thing for outdoor positioning is that SuperTag can operate in that hub mode. And so now not only does it, you know, locate itself via these technologies, but now it can locate Bluetooth tags. They inherit that location essentially. So that's a really neat feature. So we use that as well in our outdoor positioning.

Carson: Well, and I think so too, using that digital leash is super cost-effective too for customers. And we can find, we found that recently that you can get 10 or 15 SuperTags and then you can tag a thousand assets and have it as such a, you know, high volume, but an affordable price. So would you say that there are any other technologies or techniques that we haven't covered that are prominent in the location technology space?

Mark: Yeah, yeah, and yeah, there is. Like I said, there's… several niche ones. You know, some of them could get downright weird. I've seen systems that try to use the Earth’s magnetic field to position themselves.

Carson: Wow.

Mark: Yeah, it's very research papery, but I mean, I think it takes like hours and hours and hours to get a fix, so it's not terribly useful. But I think in terms of the kind of the ones that you do actually see, there's ultrasonic positioning. We ourselves have a system that uses ultrasound. That's our associate alert system. And we use that for room-level accuracy. Ultrasound, it doesn't penetrate walls. And so, you know, for that system, we added ultrasonic speakers to our beacons and ultrasonic receivers on the tags. And then they do that kind of first-tier positioning, proximity positioning, and then they connect to the beacon and request an ultrasonic pulse and listen for it. And if they get it, then they can be 100% certain that they're in the room with that beacon. So that gives us, for hotels or offices or whatever, that gives you not just room-level accuracy within a floor. It also gives you floor level accuracy because you know exactly which room you're in. But you know, it’s a neat system. I have seen other people working on ultrasonic positioning to do kind of XYZ type positioning, like we do with our phase-ranging UWB systems. And we did actually play around with that a few years ago, but Doppler ends up being a problem; Doppler shifts and things like that make it hard to receive those signals if something's moving quickly. So, you know, we went with just room-level positioning for that.

Carson: What is a Doppler shift?

Mark: So the example you always hear is if you've ever stood by the side of the road and a fire truck or an ambulance comes by and as it's approaching you, the siren, you know, the pitch gets louder and then it passes… I’m sorry, not louder, it gets higher pitched. Then when it passes, it gets lower pitched. And that's just because as it's coming towards you, it's transmitting this sound wave, but it's also approaching, and so that sound wave gets compressed. And that's basically like a frequency shift. And then the opposite happens when you go away. It stretches that signal out, and so the pitch goes lower. And the same thing actually happens to RF waves as well. Uh, you know, like satellites, GPS, for example, you can get double shifts on those signals as well. So, I mean, any type of satellite communications have to account for that because it can be really severe, especially if you're talking to lower earth satellites. Yeah, so for ultrasonic positioning, because the frequency is comparatively low, it doesn't take much to get a really significant Doppler shift to make it hard to receive reliably so, yeah, so that's one system.

Mark: Another one is a technique could be RSSI fingerprinting, which is effectively, you know, you set up beacons that are transmitting in a space. And I've seen this applied to very large environments, hospitals, things like that. And then you go around, you effectively survey the environment. And, you know, because in theory, when you're moving away from a signal source, the signal strength should roll off smoothly, but in practice it does not. You get these what are called nulls and things like that due to multipath bounces and ground reflections and stuff. And so what fingerprinting does is it effectively kind of bakes that into the cake. You come and you just measure and then say, well, here's the signals and strengths that I see at this location. And let me look in this database and oh, that looks like this location. And I think that can get pretty accurate. The issue is if that environment changes at all, now you need to update that database. And if you're in a very dynamic environment, that database will need to be refreshed frequently, which can be time-consuming and possibly expensive.

Mark: And we briefly touched on angle of arrival. That's the other kind of technique you see a lot. We did look at it. Just, we didn't like the fact that it requires these relatively large custom antenna arrays. And the processing for that is too intensive. We wouldn’t have been able to take our general approach where the tag calculates its own location. We would have had to do that elsewhere because the algorithms are too much to run on a tag. And so that would have meant having to have wide bandwidth backhaul from all of those receivers, so then you're talking about running cable, ethernet cable and installations and stuff. And, and that can get expensive fast, so we shied away from that.

Carson: Well, you've obviously seen a lot of location technology advancements, and obviously, you know, a lot more. And some of the things you're saying are way over my head. But I'm sure there's going to be a lot more advancements to come in the future. What would you say are your predictions for the future? Do you see… You know, something exceling over something else or something new coming or what? What would you say?

Mark: Yeah, I mean, I'm generally not great at predictions, but I think if I had to put my money down… I mean, you know, UWB is… I think cost is going to continue decreasing and we're going to see more integration. You know, right now it’s a separate radio – a separate chip on the tag, I should say. But I wouldn't be surprised if at some point we're going to see an integration increase. It'll get merged, maybe with that Bluetooth chipset, which would be much smaller, lower cost, you know, probably more power efficient. I think that's definitely happening, especially with Apple, as I said, getting into the UWB game and it getting put on smartphones now. It's definitely here to stay. I think the other thing is that Bluetooth is constantly updating their specs and adding new features, and I think location accuracy is obviously very much on their radar. I mean, they added AOA. I think they're going to keep improving and adding more location options and improving accuracy as well, effectively trying to keep pace with UWB, I'm sure. So obviously, we'll be poised to benefit from both of those advancements. And then kind of next gen 5G cellular, I've seen lots of talk of positioning features they’ve kind of baked into that. And so, once those kind of come online more, it's another area where we're kind of ready to jump in with both feet.

Carson: Yeah, well, Mark, thank you so much for joining me again. I know a lot of the stuff you said, again, is way over my head, but I know that there's listeners that, they've been doing this as long as you have and they have that technological mindset. So this will be very helpful for when they're, you know, making choices on the best solution for their use cases and everything of that sort. But thank you so much for joining.

Mark: Oh, thanks for having me.

Carson: If you could, tell our listeners how they could get connected with you if they have any questions about the technology.

Mark: Sure. Yeah. I mean, I'm on the website. My email address is mark@link-labs.com. So, you know, welcome any questions.

Carson: How did you get Mark? I want carson@linklabs.com.

Mark: Yeah, we did that early on. They gave everyone, I have both mark.bloechel or Mark. But I'm getting my last name, that's easier.

Carson: Alright, man. Well, thank you so much for joining.

Mark: All right, yep, thanks.

Carson: Thank you so much for listening to our episode this week. Link Labs is a leading innovator in all things Internet of Things. Link Labs offers an asset tracking solution that uses technology to improve companies' efficiencies. If you want to learn more about Link Labs and asset tracking and all the many benefits that we can provide to your company, visit our website at link-labs.com and be sure to follow us on all of our social media platforms at Instagram, Twitter, LinkedIn, subscribe to our YouTube channel, and of course, subscribe to this podcast for more episodes to come in the future. We look forward to seeing you next time, and as always, thank you so much for listening in.