Mister Beacon Episode #72
Current & Future State of Location TechnologyMay 08, 2018
Grizzly Analytics is recognized as one of the world’s leading analysts focused on location technologies and IoT, based on their technical expertise and hands on approach to evaluating vendor claims.
On this episode of Mr. Beacon, we talk with Bruce Krulwich, Grizzly’s founder, about the current landscape of location technologies, how the different offerings are evolving, and expectations of the future. We talk Bluetooth beacons, Ultra Wide Band, Ultrasound, SLAM, Machine Vision, and more.
Receive new Mr. Beacon episodes right in your inbox
The Mr. Beacon podcast is sponsored by Wiliot, scaling IoT with battery free Bluetooth.
Steve Statler 0:17
Welcome to the Mr. Beacon podcast. I am really excited to be talking to Bruce Krulwich, who is a bit of a hero of mine when I was writing the Mr. Beacon, book, beacon technologies, the Hitchhiker's Guide to the beacon system, he was one of the folks whose writings I read to try and get up to speed on a broader set of technologies than Bluetooth beacons. So Bruce, you're the chief analyst at Grizzly analytics. But you're also kind of doing quite a few other things for if your LinkedIn profile is to be believed. Your background is pretty interesting. A lot of prestigious academic institutions, Carnegie Mellon, Yale, Northwestern, and your PhD is in artificial intelligence. How did you get into the indoor location space?
Bruce Krulwich 1:13
Well, what I actually was doing when I first started Grizzly Analytics was tracking technologies as they developed in patent filings and in research and in what was coming out of the major mobile companies, and trying to understand what the trends were going to be, and using the technology background, to be able to look at a patent and look at 100 patent filings and find out what the trends were going to be. And in the middle of 2011, I believe it was, I noticed a huge uptick in patent filings that as I thought about them related to indoor location was somewhere in beacon technology, somewhere in radio technology, somewhere in motion sensing. They were in a lot of different areas. But the commonality was indoor location tracking or GPS, Lis location tracking, and 2011 is when I got to jump then on writing reports on the subject, and getting into the subject as an analyst before the trend really hit the market.
Steve Statler 2:12
And so why were you spending your time looking at patent filings?
Bruce Krulwich 2:18
Because that when I started Grizzly analytics, that was an area that the analyst, a lot of analysts are coming from a business background, coming from a technology background, I thought there was a niche to be looking at the technology trends and understanding them in those way before they got to market.
Steve Statler 2:36
Okay, that's smart. And maybe you can tell us a bit about what the focus is for grizzly analytics, because proximity location technology is such a broad area, how have you decided to focus your time as an analyst?
Bruce Krulwich 2:53
I would say the main goals and lives main focus is on the technologies is on how one technology differentiates itself from the other, what different technologies are useful for and what the implications of those technologies are.
Steve Statler 3:10
Fantastic. Well, let's talk about that. And on the show today, I'm just really interested in picking your brains about how you see the state of the indoor location ecosystem, where it's headed, interesting companies and the kind of technologies that we should be looking at, and maybe kind of give us a quick guide to some of those. And then let's end up and have you give us a preview of where you guys gonna be spending your time over the next year. So we can look out for, for for those work products, those reports that you're working on, but let's kind of start off at the beginning. And how would you sum up the state of the indoor location market at the moment?
Bruce Krulwich 3:52
Well, I have believed for years, that indoor location was going to take the world by storm. And unfortunately, in from for a number of years, when I would say that, at the end of the year, people would say well, has it happened yet? And I think it hasn't quite gotten there yet. But I think that there are a lot of things happening in the indoor location space right now. That it really is the time right now. The accuracy is getting better. There are systems that can work with different kinds of infrastructure or with no infrastructure. And I'll explain all this in a minute. And instead of focusing very exclusively on retail kind of applications, how can you find your way around a store? There's a lot more look now at other valuable areas like analytics and location and back end services where there's a lot of value coming out of indoor location in ways that are that are worth a lot more money to a lot more people.
Steve Statler 4:50
Okay, well, that's very interesting high level statements there. So there's a number of themes one is locations getting better infrastructure is becoming more affordable or less of a less of a impediment. And it sounds like we're figuring out what the use cases are, that are really going to drive adoption. let's kind of take those one at a time. Where are we in terms of the kind of accuracy that we can expect? Because I think a little bit like you, I've, I've looked at this ecosystem, and you read a whole bunch of stuff in the marketing literature about what can be achieved, but it's very rare, you actually see that in practice, and one of the things I really like about what you do, is that you I mean, you're a CTO, and so you, I think, have a fairly healthy disregard for for the marketing thing. And you tend to, like goI IoT last year, in Brussels, you actually had kind of a bake off thing where you put people through their paces. So So what are you seeing in terms of the reality of the level of accuracy that is achievable? And and who's able to deliver the those levels that are kind of key to crossing over the chasm?
Bruce Krulwich 6:09
Okay, it's a great question. The there are really two different categories, I would say, when it comes to accuracy, the bulk of the systems until now have been getting accuracy at maybe four to five meters. That means that wherever they think you are, you're within four or five meters of that place on average. So what we saw at Geo, IoT about a year ago, is a number of solutions that managed to get that down to two meters, indoors, blue lock, and Gyps tech are three companies that managed to get that down to, to around two meter accuracy. And if you think about it, two and four sound awfully close to each other. But in reality, two meters to four meters is the difference between whether you know what aisle of a supermarket you're on, if you're off by four meters, you might be on aisle eight, and think you're on out nine. But if you're within two meters, you're gonna get the aisle of the supermarket about right. So that difference is really key. The second thing that we're seeing is a small number of companies who are coming out with much better accuracy, you've got a systems like Koopa, coming out where they can get you down to about 20 centimeters, you've got some other technologies Marvell mind was also at Mobile World Congress recently. And they with their sound based technologies, were also able to get very, very accurate measurements, although not on smartphones yet. So you also have a lot of electronics based on UW B chips from decawave. And other companies that are getting down to that 20 centimeter range, although again, not yet on smartphones, hope is the only one that has that on smartphones. But when you get down to that level of accuracy, all of a sudden, you're in a different realm even more so you can tell exactly not only which aisle you're on, but whether you're standing in front of the pretzels or the potato chips, and then you can do a whole lot more, that takes a little bit more work to get that.
Steve Statler 8:02
Are you seeing people, companies that actually want that level of accuracy? Because I think that the danger for people like you and me is we kind of were fascinated by the technology. And you know, we want more accuracy. And you've done a good job in terms of framing accuracy in the context of retail, but I kind of go through these waves of excitement and disillusionment as I see these new technologies, the ultra wave, ultra wideband and angle of arrival technologies and ultrasound. And then I kind of look at what actually people need. And sometimes it's actually that they don't need that level of accuracy. So are people willing to pay for this stuff?
Bruce Krulwich 8:48
I think the end I think one thing we're seeing in the market right now is that the kinds of applications people are using this for is broadening. So you're right if you're in a mall, and you want to know whether you're in a coffee shop, or in a sneaker shop, too, because you want to you want your friends to know where you are to share your location, you can be five or six meters off, and it doesn't make a difference. If you're in a supermarket, as I said, two three meters makes a big difference compared to less accurate. But there are a growing number of applications where there's a lot more value for more accuracy. For example, if you are on a factory floor and you're trying to see whether a person is entering an unsafe location, he might be a half a meter away from something dangerous and he's perfectly safe. But that half a meter or 20 centimeters would make all the difference between being hit by the forklift and not being hit by the forklift. Yes. So when you have applications like that, which is what you're seeing from companies that are bringing more accuracy to market, whether it's whether it's COPPA with their angle of arrival technology or you web technology, those kinds of systems are being used where the accuracy makes a huge difference. And those are things that that two or three meters off, and you wouldn't be able to do it wouldn't be able to operate.
Steve Statler 10:10
Yeah, I think you're absolutely right. I think there's also just an expectation management when you promise one thing and you deliver another and then it just kind of freaks people out. And I think people feel like they should be offering meter level accuracy, and very few people can, or what I've also found, as a writer of the occasional RFP is you specify one meter accuracy, because that's what everybody says. And and then you realize, well, that means the skies are darkening. Because there's so many hubs and receivers being installed in order to get this that not only does it become incredibly expensive, but just politically, anyone that's installing that infrastructure becomes a little bit embarrassed about the fact that they've got tons of cabling, and all that sort of stuff. So So let's segue on to this infrastructure bit and give the give us a sense of where you see the state of the art in terms of the infrastructure required to get the kinds of accuracy that we've been talking about.
Bruce Krulwich 11:09
I think the I think the state of the art is really moving in two completely different directions. On the one hand, you have systems like Koopas, where they can give you 1015 centimeter accuracy. But you need to use use locators that are more complicated than anything else out there than beacons that are out there. If you want that level of accuracy, you don't have you don't have that much of a choice if you want it for smartphones, so you'll go for it. If you want to meter accuracy, or three meter accuracy, a lot of the regular use beacons or Wi Fi based systems can achieve that now. But another new state of the art trend is infrastructure free solutions. If companies like at most recently, at Mobile World Congress, there was a comp, a company called system and a company called Gyps. Tech, both of which are using infrastructure free. And maybe if there are some signals around from the environment, they'll use them. But they're basically doing motion sensing and using no infrastructure, and they're getting reasonable results. They're not getting submitter, but they're getting very good results, given that they need zero infrastructure. And and they do the best they can with that. And that has, that's obviously would be usable for a whole different realm of applications. If you want to have always operating location, or you want to know where you are, wherever you go without having to install any infrastructure. That's a whole different state of the art.
Steve Statler 12:33
So what you're seeing is these some of these MWC exhibitors that you were tracking, Mobile World Congress, exhibitors, they're using the accelerometer to kind of figure out what the relative movement is, and then how are they how are they inferring this with no infrastructure seems too good to be true.
Bruce Krulwich 12:56
So the idea is like this, if you have a button, bone and smartphone, smartphones, as we know, can tell whether you build like this, like this, right? So they can tell the movement. So if you can tell your smartphone here, moves like this, and you can track the speed, the orientation and the direction where it moves, then you can tell if you knew where you were here, you can tell where you are here, given how much you've moved, right? So that Intel recently, that has tended to work reasonably well. But error has built up over time. Yeah, and several companies are starting to work with that sort of same technology, but tried to cancel out the errors using a variety of other new methods.
Steve Statler 13:39
But you got to have some kind of approach going. So that's kind of changes in relative location, but you have to establish your position at some point. How are they doing that?
Bruce Krulwich 13:50
Sure. Some of them are starting out with GPS when you're outside before you walk in the door. Okay, some of them are trying to catch GPS, when you walk near a window, if they can catch a little bit of GPS, they'll use that. And many of them are also using what you'd call signals of opportunity, which is if you're walking around the mall sent by based in tracking location based on the movement, and all of a sudden you see the Starbucks Wi Fi, then you know you're near there and you can make any corrections you have to make.
Steve Statler 14:20
And I should say we are getting a little bit of break up on the line. But to my mind the the the level of insight you're offering overcomes any issues strain that may be required to to interpret the occasional blip back to what you were saying this or these technologies, do they require the app to be in the foreground or can they work in the background?
Bruce Krulwich 14:46
Many of them require an app to be in the foreground especially earlier on in you know, until the past year or so, more and more complex solutions are trying to develop systems that can run in the background. Because for exactly that reason that people tend to put their phones in their pockets and not hold them all the time. Yeah. And they want to be able to track the location passively. This also has to do with a big move recently, not away from but in addition to user facing applications, like maps, and things like that, we're seeing a lot more work on analytics, where if a company can know roughly where you are all the time, they can analyze that it sounds big brother. But in the end, they're analyzing that in order to improve the service that people get. So they know, people that walk by the tend to walk in the following ways and they can offer promotions. When you do, then it might be worth worth everyone's while to know where you're walking.
Steve Statler 15:48
And have you had a sense? So these very accelerometer orientated tool sets? Is there kind of a trade off in terms of battery life? Because accelerometers tend to be quite hungry, don't they, in terms of power?
Bruce Krulwich 16:00
They, they do. And a lot of people are working on that. But that is that is still a problem. Yeah. And then it's one that they have to overcome. Over time, the better the chips are getting, that's we're getting better and better chips in our phones than that need for battery life is going down because more work is done by the chips and not by the CPU.
Steve Statler 16:23
While you're talking about the chips, let's talk a little bit about the operating systems as well. So you know, what Apple do and what Google do a super important in this space is they own the operating systems, how would you what what are your observations on their direction in terms of indoor location technologies and what you get from the operating system, how's that changed over the last few years.
Bruce Krulwich 16:47
Apple has done a lot of made a lot of acquisitions in this space, from Wi Fi slam onto several companies. More recently, they've consistently been looking at what are called slam technologies, which are technologies that can learn about a place you walk into and basically learn whatever they need to know in a new site to be able to provide indoor location in that site. And they've invested in that tremendously. And it's not 100% clear whether that's already right with the technology that they've acquired is already running in iOS now, or will be soon. But they're clearly getting ready to be able to have iPhones taken into any new site and learn what they have to learn from that. Google has, has not to my knowledge, put a lot of investment work for sure into slam technologies, their focus has always been learning about all the radio devices that are out there and approximately where they are. So they have the biggest one of the biggest databases of access points and Bluetooth devices that are fixed. And they traditionally have always used that to do localization. They also did a lot of work recently with Project Tango, which is more visual based and camera based. But it's not clear yet whether that's actually coming to market or how that will be coming to market.
Steve Statler 18:13
Yeah, that's interesting. Yeah, I think it's funny that Apple kind of lit the fuse on this, on this beacon thing. They weren't the first to do it. But they certainly poured gasoline on this fire to mix my metaphors. But they don't appear to use beacons, they don't have a beacon database. And they don't appear to use them in their maps product. But but Google does. And the way I think about it is Well, let's think about how these guys make money. And for Google, they make money with advertising. And so I think beacons are a perfect fit for them. Because they can do, they can do attribution, and they can, they can measure, basically, whether you went to the location that was advertised. And so it really makes sense for them to have that very precise repository of beacons, whereas I think Apple is less, less worried about that they don't have an advertising business to speak of. And it's kind of more about, can I just get good responsiveness from from the map and think that drives some of some of this?
Bruce Krulwich 19:19
I might, I might add to that, that, as you said, Google's main business is advertising and is in whatever conserve that advertising. Apple would love it. If for any app that can run only on an iPhone, justifies whatever Apple does, Apple makes most of its money off of its device sales. So if there is going to be a couple of high end apps that will do something new and innovative on an iPhone that can't be done on Android. Then Apple calls that a win because that if that can motivate 1% More iPhone sales, then that works for them. So their big push seems to be how Look in their operating system enable apps to do more things that are unique.
Steve Statler 20:06
Yeah, that makes sense. So let's get back to the technologies piece and the use cases. And you you mentioned some of these infrastructure free solutions, which are great if what you're trying to do is tracking smartphone. But as you alluded to earlier, there's a lot of action going on, that isn't involved with tracking consumers in retail environments, the whole real time location systems where you may be tracking tags that are on pallets, or even going forward tags that are built into the products themselves that are being sold and the and the packaging. What are the technologies that you see as being kind of the most promising in in that space?
Bruce Krulwich 20:54
The technology right now, if you look at what I would call location enabled electronics, which I think is a huge new area that's only now really coming out to market, I think UW B, which is you know, largely based on chips from deco wave, although there are a few others in the space. But they dominate the market is clearly the leader in the location enabled electronics space, because they've got the technology down to one chip. So if you can get a chip into your whether it's a drone, whether it's a robot, whether it's a watch you wear while, you know, put you put on to a kid when they're running around the mall, so they don't lose them. Whether it's any one of these location enabled electronics kind of products, then that, you know, the you web technologies is very strong there. So I'll draw Yes, there is some, yes, ultra wideband is very big there because the chip implementations sorry.
Steve Statler 21:54
Now, just to explain that, so Ultra wideband is we don't have huge web radios in our, in our phones, and ultra wideband has been around for a while hasn't it deco wave is not a new company. Maybe you can kind of explain a little bit of the trajectory because, you know, I looked at this a year or so ago. And it seemed like there was really a transition going from these legacy Ultra wideband systems that were really expensive, very infrastructure heavy. And literally 10 times more expensive than the Bluetooth beacon solutions. That some of which you've you touched on early, earlier. But I saw a bunch of very young companies that were starting to adopt that technology, but it just didn't seem to be ready. Where are we? Can you add some more color to that? And where are we on this trajectory?
Bruce Krulwich 22:50
Well, ultra wideband has been around for a long time. But it was only a couple of years ago that deco wave came out and then some other companies as well. But deco wave primarily came out with your web on a chip. So that you could basically have one chip in your moving object, one chip in some locators that are based in fixed locations around a room around a space. And that you could then track the locations very accurately, you base that with very low power, and only a very small footprint because it's only one chip. And that once that came out a couple of years ago, it then took time for that to that those chips to enter electronics. But what but we certainly have seen a big push in that area now decawave Just got as making a big push now and in the Chinese manufacturing market. So they I hope hope that means that we'll see a lot of well, you know, low priced and and you know, Chinese made electronics that are location enabled coming into the market would be very nice for the market.
Steve Statler 23:52
And what are the implications? Do you think for all the Bluetooth vendors because they obviously have a lot more than one primary vendor, they have a whole ecosystem of people that are providing chips at huge volumes. And, and of course, we've got the advent of angle of arrival, which is moving from this boutique technology that Cooper really pioneered from from from my perspective and advice that I would say is actually pretty simple and compact. How do you see the Bluetooth versus you web battle shaping up?
Bruce Krulwich 24:31
I think that you can't get around the fact that the phones we're carrying are have Bluetooth in them and they don't have your web in them. There was a push for a while and there still are people talking about getting you web chips into phones. And that certainly could happen. But for the meanwhile, anything we want to do that's going to track the mass market not or locate, provide location services to the mass market is not going to necessarily be on your web for the super mass market because they're not the web, get in phones, your web may be good for particular location nibbled electronics or things in a particular site if you have watches or tags of a particular site, but when you have when you turn around then in the Bluetooth systems can work anywhere on any phone or any tablet, and increasingly, any anything else, then the coverage of Bluetooth is unbeatable. Yes. And that's why we you know, when you see the accuracy issue also being discussed. In the end, if we want to track smartphones, we have to use the technologies that are available there.
Steve Statler 25:37
Okay, so Bluetooth has a role tracking smartphones and doing RTLs applications where there are no smartphones, for my money, there are always going to be kind of a lower cost choice just because of the economics, the volume of the of the chips. What about some of the other technologies that you look at ultrasound, and so forth? Where does that ever play?
Bruce Krulwich 26:01
They're all a lot of them are doing? Well. The one that I'm excited by in theory, although we're still waiting for it to come out more is the is what's called VLC, which is visible light communications, where you have companies like acuity that have purchased startups, they've purchased byte light a number of years ago, yeah, and these are, these are lightbulb makers, these are people that are in a not, you know, not the highest tech of industries supplying light bulbs to mass sites. And yet, they're now able to put technology in those light bulbs that can be picked up by a camera and provide very accurate positioning for the for, for those devices. Yeah, catch is it'll only work when you've got their light bulbs deployed all throughout a site. So I, once they can get out as acuity and other companies that are bringing that to market, get it, get it into the market, that could be an amazing opportunity that whenever you open up your device, open up the camera real quick, get some some readings from the light bulbs, which are overhead. And and you know where you are very precisely. But that's been slow getting to market because it requires people to replace all their lights. So it's going to be a little bit slow going into market, it may be a little bit more time till that's in the market.
Steve Statler 27:18
Yeah, I totally agree with you, I'm so excited by this technology, because because it does X, Y and Zed or Z depending on which country you're in, and it does orientation. So you can literally use your phone to try and find that kind of sub brand of aspirin or whatever you can, you can use it as that pointing device. And I think that's just going to be so cool. But I agree with you, it's been so slow, because you know, the people that are running these, I think they have a lot of challenging problems that have nothing to do with the technology, per se, it's about, well, this needs to be deployed in all of our stores, we can't have a product finder that only works in half of our stores. And then, you know, if we're going to have a second source for lighting, then you know we are we really going to get all our stuff from Osram or all our stuff from acuity or GE Lighting or whatever. And and it's it's the it's the most infrastructure one of the most infrastructure, heavy solutions. So I think it's kind of was very exciting. And it went very quiet. But my prediction is in the next year, there's suddenly going to be a burst. Because I know that there are some very big retail names that have decided they're going all in and they're getting, you know, they have huge coverage and they're going and deploying this lighting. And I think the experience will be a magical experience rather than the kind of over promised under delivering experience. But the cool thing is for those of us that do consulting is it's a great tool for some things, but not for other things. And so you got to how can I get that phone out of someone's personal pocket or we'll get the app running. So the using the camera can do that. VLC stuff, there's a lot of fun solution design that will be required to make these things successful. So anyway, that's just my rant on VLC, I agree with you. So what have we missed out? So we're you know, the the overarching question was, where is this ecosystem headed? And we've talked a bit about Google and Apple and infrastructure and the use cases moving away from retail. Is there anything else we should cover? Before we sign off?
Bruce Krulwich 29:29
The remaining question that I always have is whether we're going to ever hit a point that indoor location will run on every phone out there in the infrastructure will Apple, Google or both. If both of them would give us maybe not sub meter accuracy, maybe not two meter accuracy, but better than they're giving now. If they would each give be able to give four or five meter accuracy everywhere. Then that would be a game changer altogether. And A it would it you know, it would it would move, move the bar away from the localization and onto the services for all the other companies in the space. But B would open up social. You know, we're talking about technology on mobile. And we haven't yet in this in this podcast today referred to social, because social doesn't necessarily factor in if it only works in some stores or some places. But if the technology could be everywhere, even if it wasn't so accurate, even if it was 10 meter accurate, then if I could always know that I'm in Starbucks, and not pizza, or pizza, not Starbucks. Or if I could always know what you know which section of the department store I'm in so that I could tell my friends where I am. And I could meet some friends there. And I could I could get alerts when my friend is one aisle over in the store, then all of a sudden, indoor location becomes as well as mobile devices have in the past couple of years.
Steve Statler 30:59
Now, I think that's a great point. And I am so looking for and there's a few people that have been working on this app, the the location, proximity app that links with LinkedIn, so that I actually can remember the name of the person that I just see a few meters away at a conference, that would be so cool. Okay, well, great insights, where can we get more what what, what's grizzly analytics going to be doing this year? And what should we be looking out for in terms of publications from your firm?
Bruce Krulwich 31:29
What is always coming out with new articles, and people can look at our website to track this. A new report we're coming out with we've never done before, is on mergers and acquisitions in the space of indoor location. We'll have the report coming out in about a month. So I think this new report on mergers and acquisitions in the indoor location space reflects the maturity of the market that people are interested in not only in the technologies, but also in where they're going, and how the startup companies bringing them out, are being acquired by entering into the major mobile companies. So that's a new report we have coming out that we're very excited by.
Steve Statler 32:07
I think that's so onpoint because we are crossing the chasm. And part of what happens is a lot of the companies disappear, and some of them don't have anything to offer and they just go bust but many of them have some great elements and they just need to be integrated to to other companies, or maybe they're just going great guns and there's more value from those acquisitions. So I think that's gonna be a fascinating report. Can't wait to see it. All right, Bruce. Well, thanks very much. It's been a real privilege to have some of your time on on the on the show. And I've read your stuff forever. And we met briefly in a geo IoT and it's been great to spend a little bit more time picking your brains for the benefit of all of the viewers and listeners. Thanks a lot.
Bruce Krulwich 32:52
It's my pleasure. I love your podcast and it's great to be on one. Take care.