You're listening to episode 741 of A Very Spatial Podcast, July 14th, 2024.
Hello and welcome to A Very Spatial Podcast. I'm Jesse.
I'm Sue.
I'm Barb.
And this is Frank.
And this week, we are approaching the Esri UC, so it's kind of weird, like there's a lot of stuff that's going to be talked about tomorrow. That we'll talk about in like a week and a half to two weeks, but until then we'll talk about other things.
Well, plus a lot of things that they were going to talk about tomorrow, things that we kind of talked about last episode, too. So
did we, I guess we did
the June update and, you know, honestly, the June update covers an awful lot of what they're going to talk about.
I assume there'll be new stuff. of some sort in the first couple of hours.
Well, we have a couple of people from West Virginia that are there, and hopefully we'll get we're all on a text chain. So hopefully we'll get updates of neat things.
And I'm attending virtually, and I sent in something for the plenary, so I'm going to be watching to see.
Ah, the plenary.
That's right.
And then, of course, first up this week, though, about the general geospatial industry. IDC's most recent market forecast for the geospatial industry has been released. And of course, this is the thing that people pay to get money and access to. So we can't get access to it cause we're not going to give them money, but there's lots of people who are saying, Hey, give us money to tell you what this thing said. And so we linked to one of those in the show notes.
And basically it's, it's kind of saying that between. roughly 2021 and 2030, there's going to be an 18. 8 percent compound annual growth rate for the geospatial analytics market, which sounds about right. I think, right.
It's pretty aggressive growth, but I mean, it makes sense.
Well, in this, they're kind of pushing things like the AI, AI based GIS solutions, which we're starting to see a little bit of adoption, smart cities urban planning type of things. And of course, with the push from rural to urban you know, we're, we're continuing to see more and more of that continuing response to things like health and monitoring the world for those type of things, movement into AR, VR, IOT.
So you know, not necessarily new things that they're reporting on and saying are having an impact and the growth over the next, you Six years, but still things that are, are having an impact most likely.
Yeah. That was one I had a, a well in my mind. Right. It's always been a question for us was like, oh, when is the ar vr going to be the great thing? 'cause it's, it's not happened yet. I mean, it's like, there's, there's examples of stuff like that, but to have an effect on, on the size of the industry and stuff like that, I don't think it's, I mean, we've been saying that for a while. It's just still not. to me,
it's a hardware thing.
Well, as we've said, right, the, the reason for it, but I still, I would be conservative on that having a huge impact.
Yeah. It's a hardware thing. Yeah. So until the hardware is there, it's not going to have as much of an impact.
Well, Narf and the killer app, as it were like application in that sense, not necessarily software, but the thing to use it for. That's what
requires the hardware to exist first. So there's not going to be a killer app until there's hardware. There's no hardware until there's a killer app. People are trying to build the hardware. It's just not there yet.
I feel like that number is low. Like I know it's accurate, but I also feel, you know, partially because how embedded geospatial is in everything that that number actually, when I saw it, I thought I would have expected higher. Considering everything that's going on.
I mean, that's, that's the thing, right? It's a projection. So they it's, it's market guidance. It's not, you know, reality somehow people make millions of dollars. Off of things that aren't based on reality, but that's okay.
Well, I mean, like in the total number, right to 209 point something billion dollars by 2030. So and then, and then the next thing that you think about, right, or maybe, maybe only olds like us think about that. It's like, Oh my God, that's only six years away. So to grow from its present size or from its size in 2021 to that's actually is a pretty big leap.
So just to give a context in 2020, the market value was 60. 94 or 61 billion. So that's a fairly large growth from 2020 to 2030. So we'll, we'll see.
Well, well, the really interesting question here is workforce growth, right? Which is a little more nebulous to figure out exactly. What that looks like is interesting because I see on the subreddit for GIS, a lot of questions about, you know, it's obviously students who are about to graduate or recently graduating looking for their first job and what's this industry look like? What's the growth look like? What's the future? What's the market? Where do I get jobs? All that kind of questions.
And I think that that's the more important thing on the, on the street, so to speak because, you know, we could take, we talked about the hardware thing, but we could take all that money and get eaten up by, you know, hardware. And then it's, you know, electrical engineers, computer engineers and computer scientists making all the money off of all that stuff. So it's, that's a really interesting thing from our industry standpoint.
Yeah, I mean, it's one of those things that whenever we talk about these market projections, they actually, you know, from a market perspective, the, the labor force, AKA the job market as you're worried about is most of us are worried about. I'm going to start getting off onto a grouchy tangent if we keep going on that one. So.
Fair enough.
Let's talk about other grumpy things, like the thing that Frank texted me about during the week. The, the Reddit post he sent a screencap of.
Well, in fairness, it actually wasn't the, that wasn't the thing that I thought was funny. It was the response to that. I was really texting you. And what was that?
Well, I guess,
say what it is and then say the response. Or that the other part. One of the things that Google Maps, it made the, the rounds that Google Maps is going to add pop up ads during navigation to the Google Maps, which means that the way it was depicted in the original Reddit post that I sent to Jesse.
The link to the article, at least, is that you'll be driving around and it would say, Oh, you know, it's kind of like the minority report for those who remember the movie situation where you're driving around and goes, Oh, hey, by the way, the gaps got the pants you like buying on sale and you haven't bought in a while. You here's the directions on how to get to the gap, even though you're going, I'm going to the doctor, leave me alone.
I don't care how to get to the, that was the, the way it was depicted and that's not going to happen. Google is not doing that. I'm not exactly sure why that. Picked up what that
I think that it was probably a screen grab from somebody who was doing like a routing plan while they not when they were doing navigation. So the link in the show notes, you know, goes to, they're not rolling out ads during navigation, but that doesn't mean they're not creating pop ups while you're planning and looking at the Google Maps page.
Well, I mean, you know, very much. So they show, you know, this gets into whole questions about Google Maps that. When you're looking at Google maps, it shows you a bunch of places and you go, Oh, these are interesting places. Like when I went to Chicago, like here's a bunch of places that you might find interesting. And I was looking for a place to eat and all the really interesting places that I wanted to eat weren't shown on Google maps. They were showing me things like.
You know, Chipotle and Starbucks. And I was like, I don't, I don't want any, I get that home. I don't need, that's not what I want to eat. I want to eat local, interesting cuisine. So that type of thing can happen and does happen within Google maps, but it's not why you're driving around for goodness sake. That's not. makes no sense whatsoever.
Yeah. What's interesting to me is that when you watch the original geospatial revolution video series from Penn state, that's on YouTube, and that was more than 20 years ago, it happens there.
Yeah. So the response to the Reddit thing that I, that I really sent to Jesse, that was funny was somebody had said, where do you even go to buy paper maps anyway? And I was just like, Oh God, it's just was a little disheartening to me that some, the kids basically were trying to figure out how you do the whole paper map thing. And I was like, Lord, we failed as an industry to not even make that. Like people don't even know where to start to look for it.
Go to a rest area, go to a bookstore, go to a grocery, not starting at grocery, but a gas station
hit print on your Google, you know, Chrome, so whatever.
Frank. I feel there's this discussion we need to have about who owns printers.
That's true. Cause I don't, but I'm just saying there are options
or other apps you can use.
There's that too.
Barb, you want to come for this one?
The Google maps is rolling out a speedometer and speed limits on the iPhone and CarPlay globally. And this was honestly, it's the way the news got out there that TechCrunch had someone telling them they saw this in India. And then they just basically called up Google maps and said, Hey, we saw this feature, is this true? And they went, Oh yeah, we're going to be rolling this out. And it's something that's already been rolled out by Android a few years ago.
But this will be being rolled out for iPhone users.
In case anybody's wondering why I wanted Barb to cover that. It's because she's the only one of us who doesn't drive. It just amused me.
Well, I'm just curious in your new vehicle. Do you have, like, does it have CarPlay and stuff like that? Yeah. Does it use Apple's maps mostly?
If I, yeah, if I'm doing CarPlay, just, you know, it's, it's whatever app I'm using.
Yeah. So my new, my new car has a heads up display, which is cool and stuff like that. And it'll have in the heads up display, it has. Your speed and then it reads the, the speed by actually sensing the signs and reading the number off of it. And then it's also got turn directions, but only if I use Apple maps. So I don't, I don't notice this exist having this or the failure to have this because my old car didn't have car play or anything like that.
So I didn't like use it for directions and more often than not, I would use barber for directions. I would say, you get out your phone and you figure out how to get to this place first. Can I get to, and then she would tell me what it was, but on a new car, I have CarPlay, but it uses Apple maps. So I, I stopped using Google maps for almost everything. I actually, it was really funny because I pulled out back to the earlier article.
I pulled out Google Maps when I was walking around Chicago and I was like, Oh, why don't I just use Apple Maps? Cause that's what I'm familiar with now. And, and I'm here to tell you for those who may have be skeptical, Apple Maps is really good. Just it's good now.
And it also has walking directions. Do you have an Apple watch? I forget.
I do have an Apple watch. And I'll give you a turn
while you're walking.
The only problem I, yes, and it does, but I also tell it to go away because like when I was in Chicago and most cities that I'm in, it's usually a British setup where I'm walking around. And so. It just so happened for reasons that had nothing to do with anything, I just went a slightly, I went a different direction because I was going either had to go up two blocks and over one or over one and up two blocks and it kept trying to do the other way and I was like, go away. I know where I'm going.
It's not that hard. It didn't like that. I was changing the the, the route.
But yeah, so all of them support their, their own thing. Speaking of supporting your own thing, we of course are an industry and we want our industry to share their thing. So the folks over at geo hipster. I want you to share your work by submitting your 2025 calendar submissions. They're taking submissions through October 15th.
If you have an awesome map, an awesome project that you're working on that you want to share as part of your work that you're doing, you just want to get it out there, send it over to the folks at Geohipster and of course you can always find them over at geohipster. com for all the things that they do. Throughout the year. It's not just the calendar.
Yes. And if you're, if your map is selected for the calendar, you'll get a free copy.
You can also be a patron, which I was looking at. I didn't know that, but I'm going to be submitting something. I've been trying to like support the people that are out there doing stuff. Even though I know that sometimes you run that thing of, I've made a map, but it's about a very serious topic. And I don't know if anyone wants to have this come up for that month. It's always, you know, back and forth between information versus something that's going to be. Inspiring when you see it,
put it in there. You never know. It's up to them to decide.
That's right.
And the criteria is it should be hip, but it's up to you to find what hip means. And geo, of course. Yeah. Oh, it's gotta be a map. It's gotta be a map and it has to be hip, which I'm curious if the kids these days. I bet they don't.
Yes.
And now Frank's made you hipster feel old.
Yeah.
To be fair the National Spatial Reference System is getting updated, whether you think it is or not. This National Spatial Reference System 2022 is happening and this isn't news. It's just a reminder. We just want people to be thinking about it.
Unfortunately, I feel that it is news. And it really, really shouldn't be because what is the name of the national reference system? It's the national spatial reference system, 2022. It's two years ago. It was named for goodness sakes. This is not news, but a lot of people are taking the news
and it's not expected to actually be out there until 2026,
but there's alphas out there. They want you to know the kind of the reason I wanted to bring it up now, you know, just. Besides, it's been two months since we talked about it last time is the fact that only 11, I think maybe seven, seven or 11, seven, 11 states have actually begun to put through legislation or pass legislation regarding their state plane coordinate system. That's going to be using.
NSRS 2022. So the state plane coordinate system acronym 2022 is kind of not moving forward very quickly. Some of that might be because they've been waiting for spatial frames and things like that from the the spatial reference system. But it's just, you know, kind of keep an eye on it. Our state, I don't think either of our states have their state plane. legislation done yet. North Carolina does. So where I was does have it done.
And a few, it was like six other states or 10 other states, depending on which one of the two numbers it is. So it's just something to be aware of, but in general then, yeah, should be aware of the national spatial reference system. Any thoughts?
Well, I, I should say I was kind of joking, but to be fair, like so much testing and, and work has to go into this change because of how important it is for both its accuracy and precision. So I forgot what I was going to say at the end. Sorry, I forgot the second half of that.
They have the conversion tool that will be linked to that that's in the, the show notes, but also NOAA has a whole, whole lot of really good videos out there online. If you're looking for more information or you're trying to explain this to someone who's in the legislature or anyone else that you have to speak to, who might not understand why you need to change, you know, start shifting so quickly or seemingly quickly, even though this has been in the works for a long time.
NGS website. Link in the show notes is to a blog post from earlier this year from Esri about how to do the transformations in ArcGIS and what's available now. So that's all that's being linked to here.
I know that I teach a lot of students that are in surveying and we've been talking about this for a while. A long time, as long as they've been in, in, you know, my minor and as long as they've been in their program this is something we want to make sure that they know so that when they go out, they can talk to other people about it.
Yeah. So I jokingly say that some elements of our state are now just agreeing to move to NAD 83 and that's a joke, but there's a little, little element of truth to it, particularly on the surveyor side of things. So this is something that. I feel that we're just not it's not something that people quite have come to grips with the implications of it. Like Sue said, there's a lot of stuff that has to go on to make this happen.
Well, and they've just sort of said, I'm not going to think about it until I have no other choice. I kind of feel like. This is not dissimilar in a very different way, but not dissimilar from the move from ArcMap to ArcGIS Pro. It's not until you just yank it out of people's hands and go, you can't do that anymore.
They're going to go, oh crap, I need to get up to speed, which is unfortunate given that a 2022 datum isn't rolled out for four more years and people are still going, I'll deal with it in a couple years. Now's the time people don't be dealing with in a couple of years. Get your head around it now
to be fair, though, whenever we started in W in West Virginia I think both of the tech center and in general in terms of stuff we were doing. And to be fair, this was 24 years ago. We were doing a lot of conversion from 27 to 83. And so
you were excited to do the conversion. Yeah.
And so, you know, it, it's not like everybody's going to jump on to 2022. immediately, but just, we do want people to be thinking about it, especially since, you know, there'll be some conversion steps that you will need to do if you want to get your data. Into 2022. And I'm assuming that, you know, as we look more and more at the online back ends, they're going to want it to be in 2022. Once that's once they move into it. Now, that's a whole other question of, you know, the time frame for that.
But
it really is going to kick in when the federal government goes. Oh, if you're submitting data us, it has to be this. And then that's when your Esri's and your Autodesk and all those companies are going to go, Oh, well, we do a lot of federal work. So therefore we're going to make you do that for all of our products. So that's really when it gets momentum and it gets that, that I hate to use the word, but synergies happen and it becomes a requirement.
And it'd be a lot easier, I think, if people start that process earlier and get their head around it earlier, rather than wait until. You know, you've got to submit for a grant or you've got to submit for a federal project or to get federal funding, you have to do this. I go, Oh, I have to learn all that stuff now. Start early people.
Yeah. Start learning now, even though you can't implement yet, because it's still an alpha, we want you to, to be mindful of it and be thinking about it and learn where you can. And that brings us to our last item of the news.
The OGC has is requesting public comment on the indoor GML version 2. 0 conceptual model. So the open, they started the opening for the comments of July 9th and it's closed August 15th at basically midnight. You've got a month ish. To a little less, but to, to comment on this so like, you know, as we just put out there, or was it called indoor? It's their indoor product came in the name of indoors.
There's a lot of movement on this indoor geographic space and mapping it and using it in a geospatial context and stuff like that. And I really think it's important to get involved in this as much as you can. And a conceptual model is a great place to start. The reason I'm moving to version 2. 0 as opposed to 1. 0, whatever the heck is the current one, because I think there's a recognition that You don't do everything perfectly the first time and you have to evolve this stuff.
And so it was, let's try to get this indoor space that we can't do anything with because of the limitations of technology. Now we have access to things that maybe we can do something with it. So let's do it. And then it didn't exactly work. It worked, but there wasn't a large uptick. And it had limitations and issues and problems. So I think it's great that they're going, okay, how do we rework this to make it better given the evolution of technology?
And so, you know, get involved if you can, if you've got thoughts on it, if you've got uses for it, if it's something that you touch into, or if you've said, I don't get the point of this, that's valid too, because you can, you know, use this to say, well, how do I make it? So I do see the point of it. It helps me.
And this is critical because I've worked with first responders who were, who Early innovators trying to do this. And I remember that there were a lot of misconceptions about, you know, what they need. So this is a chance for anyone that could imagine using this or being a part of this that might get overlooked because you're using it in a, you know, it's important, but yet it's underneath that iceberg of use to get involved.
Keep in mind, this is the conceptual model. So this is. The implement, not the implementation. This is, you know, what do we want this to look like? This is a logic model behind indoor GML. And so once that's solidified, then they'll move forward to part two, which is going to be the, you know, what are the actual schemas to make this happen?
And, yeah, I mean, a lot of it, like Frank was saying, is there's been growth and a lot of that growth has shown areas where there are new opportunities to build out these, these logic models, these conceptual models to support this area. Because, you know, a lot of, I think the first one relied on this kind of fusion of BIM and GIS ideas. And at that time, which I think was like, Five years ago or so.
Whenever the first one went through, those were still relatively new in terms of the mashup area. So we've, we've seen how they work a little bit better together now. And,
and how they, how they don't work. Yeah. Equally as important.
Yeah. So, you know, that's. That's the reason why we, we continue to create things and, you know, OGC has multiple versions of many different things. And so this is just another one of those that's getting an iteration.
The thing I really like is that there's a whole part in there on the semantic model, which, you know, is in my opinion, is one of the most critical pieces of all this. So when we say things like, well, you know, this is a piece of furniture, what do you mean by furniture? What does that, you know, depict it?
Is that, is a, If to be old, if you're in a mall and you've got a bolted in bench, you know, is that the same type of furniture as you've got a garbage can that can be moved out of the way, or, you know, can be moved around those types of things are kind of, it's interesting to me at least, and I think it's great that we're going to get to some definitions of those things.
And that's it for the news this week. We're talking to Ariel Seidman of hive mapper and. The first couple of seconds because of a technical issue got cut off of the interview. But we're going to start now. Basically, I'd asked him, you know, tell us a little bit about yourself and your experience. And he's kind of starting with early parts of his activities in the geospatial industry. So onto the conversation with Ariel.
I was at Yahoo. This was like 2005, 2006 timeframe. So it was still a fun time to be at Yahoo. And, you know, we had from a search perspective, roughly 30 percent market share, 35 percent Google was, you know, in that same area as well, from a mass perspective, we actually had like 90 percent market share. And what Google started to do on the map side to their credit was they started collecting massive amounts of data on their own, right?
We were effectively just licensing third party data sources. And as we, as Google Maps did that, and they had the Google Street View program, they had their own airplanes, they had a whole variety of data collection platforms. As they start to do that, that's when, you know, ultimately we lost a massive amount of market share to them. And it was really an eye opening experience. And I had our front row seed into a how costly it was to build.
These data collection platforms and to maintain them as well, right? Like, it's not just like, okay, you map the world once. It's like, you have to consistently up the updating the map, which is incredibly challenging. And, you know, that was really when I said to myself, okay, well, if the world is headed to this place where it is crazy expensive to be collecting and processing all this data, then maps are really just going to be owned by, like, 1 or 2 companies.
And so kind of just been on this journey of trying to figure out is like, how do you both crowdsource, right? Kind of like what we're doing the collection of the map, but also maintain the quality, right? Like, a lot of the crowdsource maps historically have done, in my view, a very poor job of maintaining quality while they crowdsourced. And so what that means is like, just the number. And so the likelihood of use cases and applications for this data starts to dramatically decrease.
Because you've kind of let the core issue go, which is quality. As we talk a little bit more about how HiveMapper solved that problem, it was not been like a straight path. I just been like a lot of trial and error through this over the years to get to kind of where we're at right now.
So let's go ahead and make that jump then. So what is HiveMapper? HiveMapper?
Yeah, so we're building this global map and you buy this built this device. It looks and smells like every other dash cam in the world and it performs the function of a dash cam. But it also has really important critical technology for the purposes of building a map. Right? So it has stereo depth camera so that we can properly position objects. It has professional grade GNSS or GPS modules so that we can properly position the device in the real world.
It has a tremendous amount of compute on board the device so that we can actually take all the imagery that we're, that we're collecting and then process it on the device, right? So that we can detect speed limit signs, stop signs, lane definitions, all that kind of information. And it's also connected, right? So that means it's like, it has a wireless connection where it can automatically upload this data in real time.
So there's these drivers, there's, I dunno, 15,000 of them roughly all over the world, driving around and just going about their day. Like we don't ask people to drive for the purposes of maps, but we say if you are driving, then we would love to be there alongside you as part of your journey. So, so that's one mode of contribution. There's another mode of contribution, which we refer to as AI trainers, and these are people who are doing data labeling as well as map quality assurance.
You know, and they're playing these like little games that we create for the purposes of doing the data labeling for the map. A. I like speed. Hey, we think there's a 35 mile per speed limit sign. Is that correct? Or, you know, we think this is you know, this sign is position positioned here. Is that correct? All that kind of all that type of quality assurance. And there's roughly about 30, 000 of them all over the world. And so that's on the supply side, right? That's how we build the map.
And then we're targeting three large segments of the mapping industry. So one is the B2B customers, right? So if you never use Google maps on your personal phone or computer, you're still interacting with Google maps through a variety of companies, right? Whether it be Gojek or FedEx or UPS, et cetera, many of those different types of organizations. Utilize Google Maps data in a variety of ways through their API products.
The second category is what's commonly referred to as just self driving as a category as a whole, everything from like a ADAS product, like Ford Blue Cruise all the way up to like robo taxis. So that's the second category that that's, that's a quite large market, very interesting market because it's still growing fairly rapidly. And then the third is consumer navigation.
We're not yet competitive there, but over time, I think we'll, we'll definitely introduce something that I think is going to be pretty compelling in the consumer navigation side.
And so then you also have within the B2B, you have the various assortment of, Technologies that both of course take advantage of the collection capabilities of these people who are helping to crowdsource the data, but also to create products. So you can talk a little bit about the various APIs and Scout and Burst that you have as well.
Yeah. So there's, there's API data products. So I would, I'd categorize, you know, what you see in the B2B side today, what we have today is primarily data products, right? You know, full blown products, like turn by turn navigation, all that kind of stuff that will come a little bit later. And so in many of the organizations that we.
That that utilize our data products today, they're effectively compliment using our data to compliment existing data sources that they already use or to get better, fresher data in specific regions and specific locations over time. There will be other data products like place search or like, turn by turn direction or all these other commonly use map API products. That that's what that's 1 day.
So we have a data product, which is basically map imagery and the data product, which is map features map features is things like speed limit sign stop signs, turn restriction signs, highway exit signs. We have another product for construction work zones as well. So that's more dynamic in nature than something like speed limit signs. And then we have a tool that you can use called scout and scout is basically location monitoring. Right? So, hey, I want to monitor these 500 locations.
So, if you're like a city municipality, and you're like, let's say, monitoring bus stops, right? Or you're doing an audit of every single fire hydrant in your city. That's a great tool, right? Or you're trying to find, you know, what are all the, like, let's say, trucking companies are very interested in, like, the heights of underpasses and tunnels and all that kind of stuff.
And so the track, all those different types of locations to make sure that their vehicle can fit in all these different challenging environments. And so scout is just a really easy to use tool where you just say, okay, here's my whatever 500 locations or 10, 000 locations, whatever it is. And I can track and I can have an analyst kind of do that and manage and track those locations.
Burst is a tool that basically says, hey, here's This location over here or these locations, I want to make sure that they get mapped much faster or refresh map much faster. And so you basically pay like a little bit of incremental bounty to, like, make sure that they get mapped a little bit faster.
And so kind of going from end to end, you have the crowdsource portion of it that, you know, invites people to take advantage of the hardware to be able to help collect this data. Then building out these data sets using the imagery and GNSS data that's coming along with those stereo pairs to, you know, kind of build out the world. And, you know, if people go over to HiveMapper and look at where you have data now, there are densities in certain areas.
But I was surprised, actually, I was recently in rural Japan, and I was surprised how much data actually had been collected in a place like that. So, you know, you do have a wealth of data, not just in the United States, but around the world.
Yeah, I think we're at like 14, like a little, you know, a little under 14 million unique road kilometers, which is like 23 or 24 percent of the global road network. You know, there's not a lot of mapping, quite frankly, that's happening in Russia and China. So if you remove those 2, which, you know, there's a lot of roads in both Russia and China, just given how big they are in terms of geographic you know, land. So, if you look outside of those areas, I think we're doing very well, right?
Like, if you look at East Asia, Japan, the Philippines, Korea, Malaysia, Australia, we're doing very well. And if you look at the EU and the United States and Canada. You know, that, that's really the bulk of our coverage today. You know, I would like to see us get a little bit stronger in Mexico and South America. And so that's definitely going to be a focus of ours, you know, next year.
And so you have, you have this variety of perspectives and how there's conversations here about crowdsourcing, about engineering, about the push into AI and taking. Data pulls from the various imagery. So it's, it's kind of an open target, but let's start with the, the question of kind of the ethos and kind of what your goal as hive mapper is to, you know, Get to these data products. You know, you talked beyond the B to B to the other two areas as well.
That I didn't really ask questions about, but you know, how does all of this kind of tied together with what hive mapper is, is aiming for?
Well, if you think about what we're aiming for, you know, today, it's basically like the vast majority of people in the world use Google Maps and ways, right? Like, I mean, like 2. 2 billion people use Google Maps every single month, every single week. And like two, you know, a couple hundred million people use Waze, right? And obviously Waze is owned by Google. So like, it's just, you know, 2. 5 billion people use a Google navigation location. And there hasn't been a ton of innovation, right?
I mean, just like they've improved themselves. I don't want to say like they haven't improved anything. But if you just look at, like, a Google Maps experience from, like, 2012, 2013 to 2024, there's like, they look effectively the same. And the reason they look the same is because there has not been. A new large scale data collection platform getting created, right?
Like you can move stuff around from a U. I. Perspective until you're blue in the face, but the thing that actually differentiates mapping products is what is the data collection source? So we've basically gone from, you know, satellites in the digital mapping age, you know, satellites was first that was obviously really important, you know, then it was airplanes. And then it was like Google Street View program. And then it was effectively motion through your GPS traces from your phone. Right?
And that was like the ways obviously innovated on GPS traces from your phone. And that was like 2009 2010 timeframe where, like, the bulk of that innovation occurred. Right. And what we're putting forward is saying, no, we now have a new large scale data collection source called eyes. In other words, these cameras that are much better at actually understanding what's going on on a given road segment than motion.
Like, for example, if you were to say, I mean, Google obviously knows like this road segment over here, people go 40 miles an hour, then boom, all of a sudden traffic slows to 10 miles an hour. They know that. But they don't know why, right? Is it because there's flooding? Is it because there's road construction? Is it because there's police activity? Is it because there's a five car accident that won't get cleared for like three hours?
Or is it like a little, you know, fender bender accident that will get clear the next 10, 15 minutes? Those are all very different things. And because they don't really know that it's hard for them to actually provide good navigational instructions. To the driver. And so what we're saying is, you know, as more and more of these eyes get out there, we will be able to build fundamentally a better, fresher map on a global level.
Yeah. I think a lot of times from a geographic perspective, I talk about the fact that, you know, we went from the top down map and we've slowly been sinking down to that first person perspective in terms of data and creating these experiences. And this of course is just another example of moving, you know, down to that person level. Whereas we're, you know, used to be at the helicopter airplane level whenever we were talking about mapping.
And that idea of space is, is something that we're shifting to place.
Yep. I think that's, that's exactly right. I mean, there's this guy, there's this guy James, James was in Esri for a long time. And then he worked at Apple and he has this great, like, Blog post, which is like, what is the next revolution in mapping? And I think he does a fairly not fairly does an excellent job of kind of laying this out there, which is like motion effectively from your phone. GPS traces is a very, very, very coarse data source. Right. And yes, it is personal to your point.
Right. But it's, it, it doesn't really help you understand a lot of what's actually happening at road level. And, you know, the, the challenge has always been is how do you get all those eyes out there? What is the incentive mechanism? How do you scale it? And how do you scale it cost effectively? And, you know, how does that device, whatever the eyes are, the camera is You know, how does it take into account privacy?
How does it take into account like the ability to actually make sense of what it's saying? Because a lot of what it's saying is like, okay, yeah, nothing's interesting here. And that's important and valuable as well. But it needs to make those kind of split second decisions. And so that that's a lot of what we're doing and a lot of like where we think the world is headed, but it hasn't gotten there yet. Right.
No one has actually been able to do this at the kind of scale that we're now achieving.
Yeah. I mean, there's a lot to be said for things like open street map, but at the same time, there's there are limitations. Like you were saying, like, right now, they're going through a period where there's people who are going through and intentionally shifting things, knowing that. You know, it's now an underlayment for so many different apps to get their own words out there, whatever they're trying to do.
But yeah, so there's, there's this disconnect sometimes between the truly open source and open access and having that little bit of, of what's a good way of putting it, but a little bit of stability, a little bit of
I mean, there's definitely, you're right. There's definitely people who are doing that, right. They're like mislabeling. You know, cities and boundaries, right? Because exactly what you said, like, they know that a lot of people use open street map. I think the bigger problem with open. There's 2 big problems with open street map. 1 big problem is the fact that most of the edits that people are making there are wrong and not because people are being malicious or they're bad map editors.
It's simply because the satellite imagery that they're basing those changes off of are a year, sometimes 2 or 3 years old and the reason for that is the, the, they rely on the donations of various commercial satellite providers. And, like, of course, you know, satellite commercial providers are not going to give them the freshest stuff that they want to monetize. Right. And so, I mean, no one has done this.
This analysis, but I would, I would put forth that probably 30 percent of the edits that are actually happening in a given month or just long because of that reason. Right? So that's issue number 1. The 2nd issue is that the fact that humans are still doing the vast majority of edits. It's crazy, right? Like, in a world of machine vision of computer vision, you know, there's still a role for humans to be clear. We use humans as well, but the initial, Okay.
You know, interpretation of the imagery of the pixels to objects really be ought to be through machines, right? It's not a good use of human talent to be doing all this stuff manually, and they just haven't progressed. And so those are the 2 major issues that, in my view, are kind of holding it back in terms of becoming a truly fresh, large scale map data source, right? There's other issues in terms of incentive mechanisms that I think it's like, look.
You know, OpenStreetMap today is used by Microsoft, you know, Microsoft is worth 3 trillion. Are you telling me that like, and you know, they use it in a variety of different ways, like they're getting a tremendous amount of value for free. And now, to be clear, open, sorry, Microsoft isn't doing anything illegal. But I do think there's this question of like, wait, what all these people 20 years in OpenStreetMaps is incredibly valuable.
And, you know, there are some people making a lot of money off of it and there's other people who aren't making anything off of it. And I, I think that's highly problematic.
Whenever we look at this ability for people to be involved in creating the data for HiveMapper. No, that's not a question I want to ask.
Are you talking about the crypto token thing that we do? Is that where you're headed or no?
No, it wasn't. But actually that's a good, a good thing to talk about. So can you talk about that?
Yeah. Yeah. Yeah. So we use a crypto token on that basically says, look, you, you were putting work in, right? Like either you, you know, you bought this dash cam device that is mapping and you're, you know, you're, you're contributing data as a result, or you're doing, you know, data labeling training of all that type of stuff. So again, you're, you're helping us build the map in some form or fashion. And so we believe that that should be compensated in some form.
And so what we do basically say, okay, look, you're going to get it. Yeah. These honey tokens and as the value of the data set that we're all building together starts to become meaningful, you effectively share in the economics of that, right? That that's really what the crypto token represents is shared economic value. And I think that's important. I get that like, you know, crypto to us is a technology and it has very specific applications.
You know, in this case, there's a blockchain where we store all of the different rewards. So if you come in and you do X amount of work, we're You know, you get Y amount of honey tokens, and that's all open. That's all transparent. So everybody can see there's no games being played or whatever it is, but it's being done very open and transparent way. And that's 1 of the primary reasons that we use the crypto blockchain technology to accomplish that. Right?
And we do believe in this concept of shared economics. Right? And so what it does is. It says, you know, look, I love OpenStreetMap and I love what it's done, but I do think it's missing that critical element, which is what is the incentive mechanism? What is the shared economics? And how do you coordinate, you know, tens of thousands, hundreds of thousands of people? Right? And so what we say here is the honey token.
That enables us to coordinate the actions of, you know, 50, 000, 100, 000 people in a very efficient way. So look, we need more contributions here, right? Or we need less over here. You can adjust that all for the purposes of building a better map. Right? And so, okay, we, you know, there's a lot more demand that's happening. And let's say for the map in Chicago or London or Lisbon. Okay, like, then boom.
You know, we will then see more contributors and a higher quality and a higher refresh map in those parts of the world because the incentive is, is there and it's kind of natural and organic to the system, you know, versus, you know, something like open street map. There are certain, you know, little towns in Germany that are incredibly well mapped. Right? But then you go to like a Rio de Janeiro, I don't know.
You know, or Sao Paulo where, you know, there's 10, 20, 30 million people who live there. That's not very well mapped, right? Because for a variety of reasons and, you know, they haven't really figured out that like consistency in really driving coverage to those areas that have demand and kind of an organic native way.
Now, one of the things that you mentioned early on. And ties into this as well is the basically the AI trainers and you know, whenever we're talking about this move into machine learning, deep learning, one of the biggest things is having that training set there to be able to give the AI what it needs to know to be able to move forward. So how are you taking advantage of this as you're building out your new data sets?
Yeah, yeah, it's a good question. You know, there's, there's some things that like stop signs, right? This is a very simple example, like the models that we have for stop signs. It was a us. It didn't take us that long to train that, right? It's a very simple sign. It's very well understood sign. There's not like a lot of different variations of it, et cetera. So, like that 1, you know. The AI models effectively have taken over now, right?
So, like, there was obviously in the early days, some initial data labeling and training that had to be done on it. But now, like, the AI models are so good that we prefer, frankly, to trust, you know, the the AI models versus a human. Now, there's other other types of signs, right? Speed limits. There's a good example where there's a lot of variation, right? There's schools. There's okay. It's only applicable in these timeframes or like, it's only applicable to these types of vehicles.
Like, we're now even seeing dynamic speed limit signs, right? Or speed limit signs that are only there when there's construction zones, right? All that type of stuff. So the variation is fairly complicated. And, you know, the amount of data labeling and use cases and examples that you need to see is quite, it's quite significant. So there, you know, you're seeing kind of ongoing data training as it finds, you know, more and more and more edge cases.
But I imagine probably in 2 years from now, like, you will have covered the vast, vast, vast, vast majority of all the different types of speed limit signs. You know, in the U. S. or Europe or wherever it may be, and that the models will then start to take over. So that's that's and this will just continue right? That that kind of cycle that I just described there, where, like, there's upfront work that needs to require followed by, you know, a I models ultimately taking over.
And then you're on to the next thing and onto the next thing. And, you know, it will go on. There's another way that we use these trainers, which is really quality assurance. Mhm. So, once the model says, okay, I think there's a 25 mile per hour speed limit sign in the school zone. There's a human that is verifying that, right? Not every single sign, obviously, but, you know, we're pulling a statistically meaningful number of the signs that have been categorized.
By these AI models to then ensure that the AI models are keeping up to our accuracy levels and requirements. And so that's the other role that they play. Now, in terms of quality assurance, like, that will be going on forever, right? Like, I can't imagine a world where you don't have humans that are doing quality assurance on the map and the underlying data. You know, you could argue. In terms of data labeling to train other objects that right now, there's a supply demand imbalance.
In other words, the demand for labeling is, is, is just really, really high because the technology has advanced so quickly. So rapidly. But over time, I think like, you know, it will slow down a little bit, obviously there will still be a role for data labeling, but on the quality assurance side, I think that that role will always be there for humans to participate in that process.
As long as there's change, there will be need for Q and A.
Yeah, exactly. There's physical change in the world that's happening every single day. And so. you know, well said, like there will always be quality assurance.
So kind of to roll all the way back to the dash cam itself. If people are interested in finding out more about the B and that process of becoming involved in collecting data, can you talk a little bit about that?
So the B is our, is our new device. It's our, actually our, our Fourth device that we built. So we built hardware, we designed the hardware, we build the hardware. We obviously work with a contract manufacturer to like assemble it, et cetera, but you know, we've now gotten pretty good at building hardware. There is a back ordered list. That's quite long.
I think it's like 30, 000 devices or something like that for the B. So that device will go into like large scale production starting the September, October timeframe. And we fully expect that, you know, it'll start to reach production scale by like December timeframe. Look, I think that the B device is fantastic because it's really a dual purpose device. It's a great mapper range.
So if you're a mapping geek or a G. I. S. geek and you want to really like, you know, have a cutting edge piece of technology, that is a great mapping device. That is it. But it's also a great dash cam. Right? So, like, if you're You know, a lot of now insurance companies will give you a discount if you have a dash cam. Right? And you may just want it for your own personal safety reasons, right? Like, if you get into an accident or whatever it is, you want to have some evidence of what happened.
And so that's very important, obviously, to professional drivers and commercial drivers. But I think it's also, you know, just just normal drivers like me and you who want to have insurance. The comfort of knowing like, okay, if something happens I have video evidence of it, so it also plays that role as well. And so, yeah, you can check it out. Hivemapper. com and then just click on B it's right there and you can learn all about the device.
Okay. That's kind of most of my questions. Is there anything you kind of want to highlight about HiveMapper?
No, I think we covered it all. I think we, we did a good good rundown of my background and, you know, talked a little bit about the product, the markets, the customers, all that kind of stuff. And then kind of dug into each one of these different capabilities. So no, this has been great.
Okay. Well, thank you very much. And of course, I think one of the things that most of the GIS community remembers from whenever HiveMapper first rolled out was the fact that you were using hexagons and we all loved that. Yeah. Yeah. Yeah. Yeah.
Using hexagons. Yeah. I think that will probably retire the, like they worked really, really well in the beginning, but now the scale is just so massive that we may move to something else. I do, I do think they were, you know, a good learning opportunity for the, for the GIS community is like they worked in the beginning, but When you're at lower scale, but now like, you know, a given road is being covered out of 80 90 times a week in some cases, it doesn't work quite as well.
The question of representation continues to go on every day. So,
yeah, exactly. Exactly. Well, Hey man, this has been a lot of fun. I, I appreciate it.
Okay. Thank you for joining us today. On the events corner as always, we encourage you to go to events, even if we don't tell you what new ones might be Of course, if you want us to add your event to the show notes, send us an email to podcast at very spatial. com.
If you'd like to reach us individually, I can be reached at Sue at veryspatial.Com.
I'm Barb at very spatial. com.
You can reach me at Frank at very spatial. com and you can follow me on any social media as @nojopar
I'm available at kindaspatial. And of course you can find our contact information over at veryspatial.com/contacts probably
as always.
We're the folks from very spatial.
Thanks for listening,
and we'll see you in a couple weeks.
I know you can fly, got a soul in the sky above.