#29 – Kevin Andrews

#29 – Kevin Andrews

March 16th, 2026

{Music}

Stewart Walker (00:10)
Welcome to LIDAR Magazine and the LIDAR Magazine podcast series. My name’s Stewart Walker. I’m the managing editor of LIDAR Magazine. My guest today is Kevin Andrews. He’s director land products at Trimble Applanix Kevin, we’re delighted to have you on board and it’s a great pleasure to be talking to you. LIDAR Magazine is very grateful to you for finding the time.

Kevin Andrews (00:33)
Thanks for having me, Stewart. Pleasure to be here.

Stewart Walker (00:36)
Well, let me give some very quick background. Your company, which was acquired by Trimble in 2003, is well known as the industry leader in GNSS, IMU systems for sensors, carried on aircraft, helicopters, UAVs, land vehicles, such as trains and trucks, and marine craft of various types. Thanks to your colleagues giving generously of their time and to the Trimble marketing folk and the communications company that you use, Keaton Public Relations.

We’ve been able to provide quite a lot of coverage in the magazine of Trimble Applanix. So only about a year ago, for example, that my guest on episode 16 of the LIDAR Magazine podcast series was the Trimble Applanix president, Dr. Steven Woolven. You’ve pointed out though, that we’ve focused very much on the airborne side and it’s time to rectify it and talk about Trimble’s offerings for land vehicles. So let’s do that today.

I always like to learn something about my guests before plunging into the technology. So Kevin, tell us something about yourself, where you come from, your education, and the path that brought you to the world of GNSS-IMU integration.

Kevin Andrews (01:50)
Yeah, happy to. I grew up in Western Canada in BC, sort of a small town in Northern BC. ⁓ I was the sort of typical kid destined for an engineering degree, know, math and science and physics and love taking things apart. And in the 90s, I was very much into computers when not a lot of people were yet. So, you know, going off to school for engineering was kind of a written. I had chosen the city of Calgary, of all the engineering choices I had.

The city of Calgary was a place that I really liked. I had family that was there. It’s a beautiful city. All of the other universities in Western Canada were great schools to go to, but I really had chosen Calgary for the location. And I had gone to do an electrical engineering degree, ⁓ bitering in computers, big sort of computer guy, and that’s what I wanted to do. But I got into university and I did my first year of engineering school where you take a lot of common courses in a bunch of different departments and realized very quickly that I had no interest whatsoever in electrical engineering.

There was something about ⁓ using imaginary numbers to evaluate AC circuits that just, I don’t know, it rubbed me the wrong way. I didn’t like it. So, you know, towards the end of their first year, when you have to start locking in on a department, they give a presentation from each of the departments. Next thing I know, mechanical gives a presentation and civil chemical engineering, which should Calgary is all oil and gas and whatnot.

And then I got the presentation from the geomatics engineering department at the University of Calgary, which I had never really heard of pretty much before then. I mean, I was aware that they were on campus, but no idea what they did. And all they talked about was digital terrain modeling and photogrammetry and laser profiling and satellite imagery and hydrography and all these amazing technologies. And it clicked. It was what I had been looking for. ⁓ And from that point on,

Geomatics engineering sort of became my whole world. I got into land surveying. I did a summer job in my hometown for a land survey company, just doing cadastral work. ⁓ And then it was in my final year that we were asked, do kind of a grad project. And we were asked to help apply the university’s GNSS, INS software, Kingspad, to a new breed of IMUs which we now know as MEMS IMUs, but at the time were very new and not very well understood. So I got a great sort of intro into GNSS IMU technology in my undergrad at the University of Calgary.

Stewart Walker (04:19)
Already you differ from a lot of your colleagues who ⁓ have typically done double E at the University of Toronto and you’ve explained very clearly why that wasn’t for you. Nevertheless, you went to York University in Toronto for an MBA. Is that right?

Kevin Andrews (04:37)
Exactly. So, you know, I finished my engineering degree and one thing that stuck with me in my final year of engineering is that many, many engineers end up being managers. And that turned out to be very true because my first few years at a Applanix, I spent a lot of time in kind of a field applications role, customer support, talking to a lot of customers. And that got me ⁓ more and more customer experience. I spent less and less time developing and more and more time working with customers and solving problems. And that got me into kind of product management.

And when you get into product management, you realize very quickly that technology is only a small part of it. And so you really need to know ⁓ more of the business side of things. You need to understand how to evaluate customer problems and how to understand business and trends and finance and all that sort of thing. So eventually became apparent that an MBA was sort of the next step for me. I chose the Schulich School of Business at York University because it’s a 20 minute drive from a Applanix.

I was able to do a kind of evenings and weekends programs so could do it part time without having to take time off work. ⁓ And yeah, it took about four years, but I managed to finish my MBA and I’m pretty proud of that.

Stewart Walker (05:48)
Yeah, it’s interesting. I mean, we’re perhaps a generation apart, but there’s parallels in the sense that I did my undergrad at my master’s rather at New Brunswick. And of course, professors went there from there to found the school in Calgary. I had similar thinking to you. I was in product management. I realized I needed to make things more systematic and an MBA will be the answer. I also did it part time. So it’s amazing how sometimes there are parallels between different people’s and geometrics.

Kevin Andrews (06:20)
Yeah, that’s a great story.

Stewart Walker (06:22)
So when you joined the Applanix, did you answer an advertisement? Did something else happen? I mean, you’ve been there 26 years, so it must be pretty good.

Kevin Andrews (06:30)
Yeah, it’s been great. So I found a Applanix while I was at the University of Calgary. ⁓ In doing that final grad project, I’d become aware of a Planix. You’re doing a lot of research into GNSS, INS. And a Planix at the time was one of the few kind of commercial companies investing in it. They were considered the top in terms of product. And then in the end of my grad year, the of my final year, ⁓ Geomatics Engineering Program posts its own job fair.

Basically, there’s only like 25 graduates from the program. We managed to host our own job fair. Most of the companies that come out to advertise are survey companies that work in the oil and gas up in Northern Alberta. But there are a few technology companies that come out as well. And one of them was a Applanix and I met my then boss who, you know, kind of met me during the day. But at the end of the day, they send everybody off to the pub for a bit of a social event. I sat down at his table and said, hey, I need a job. And he kind of looked at me over and just sort of went, okay.

It turns out I was the only one kind of brave enough to go up to him and talk to him. And his main criteria was, are you going to be able to hold your own in front of customers in strange situations? So I kind of met his criteria. I tried to hold out for ⁓ a job interview. I’d never even been to Toronto before and they were going to expect me to fly out to Toronto and try to hold out for at least an in-person job interview. It’s like, no, we see what we need to know. Here’s the job offer. Fine. So I took it, never looked back. Been very happy.

Stewart Walker (07:54)
That’s great. I think you’re being too modest. I think the fact that you were finishing a degree in Calgary might have been a factor.

Kevin Andrews (08:02)
Yeah, it helped that I sort of fit the bill quite well, for sure.

Stewart Walker (08:08)
Now for a word from our sponsor, LAStools.

—

The LIDAR Magazine Podcast is brought to you by rapidlasso. Our LAStools software suite offers the fastest and most memory efficient solution for batch-scripted multi-core lidar processing. Watch as we turn billions of lidar points into useful products at blazing speeds with impossibly low memory requirements. For seamless processing of the largest datasets, we also offer our BLAST extension. Visit rapidlasso.de for details.

—

Stewart Walker (08:43)
So Kevin, having learned a lot about yourself, let’s move on to your technology. So in order to give the discussion some context, and as you’ve rightly said, the magazines tended to focus on the airborne side. Could you explain how the challenges of land systems in terms of GNSS IMU positioning differ from those of airborne systems?

Kevin Andrews (09:06)
Yeah, of course. It’s definitely a big topic. The biggest one is obviously the environment. Airborne systems, wide open skies, hopefully, otherwise something’s gone horribly wrong. ⁓ Good sort of clean motion. It’s almost ideal conditions for collecting GNSS and IMU data. ⁓ Land is very much the opposite. You’re in urban canyons, you’re under tree canopy, you’re in and out of tunnels, you’re in and out of parking structures, you’ve got overhead wires, you’ve got overhead trains.

The GNSS environment is terrible. And I always sort of say that the only thing worse than no GPS is bad GPS, because it tends to give you a confident but wrong answer. And so that is the biggest challenge is learning to work in that GNSS challenged environment.

We have to do quite a bit to modify and customize the solution to work in that environment. We can take advantage of the fact that we’re attached to a vehicle that’s on the ground. And so there’s some trips you can do around motion and physics of that. There’s room for more antennas, more sensing equipment. There’s quite a bit you can do. But the environment is by far the biggest difference between us and the airborne environment.

Stewart Walker (10:21)
know that you can’t ⁓ disclose any sensitive company data, but could you give us some general idea of how Trimble Applanix’ business is divided up between airborne, land, marine, and any other application areas?

Kevin Andrews (10:36)
Yeah, so those are definitely our big three, land, air and marine. Most of the products are fairly similar. So we sell, you know, reference material for remote sensing, whether it’s air, land or on boats. Over the years, there’s been a bit of ebb and flow, but I think on average, the three business groups have been pretty equal.

The hydrography business that the Marine Department works for is a very alive and strong business group for us. Airborne has traditionally been a great business for us as well. And land is doing great. We’re ⁓ responding to new requests and new technologies like autonomy and mapping and sensing and ground penetrating radar. There’s a million in one applications that need a land system. So it really varies, but I think over the year, it probably works out to be about equal.

Stewart Walker (11:26)
Okay, so do you then segment the land systems market in some way?

Kevin Andrews (11:32)
Yeah, we do. So our core market, our bread and butter is mobile mapping. ⁓ So these are the truck-mounted mapping systems that are either doing very kind of low-accuracy camera mapping of a Street View kind of application, Google Street View or Apple Maps, right up to high-accuracy engineering level light letter scanning where you can do analysis of bridges and decay of structures and surfaces of buildings and that sort of thing.

So that’s our core market. But a number of years ago, we got into autonomy due to the DARPA urban challenge and grand challenge years. So we’ve had a big segment in the autonomy business. And it differs quite a bit. The mapping business is mostly a post-processed, collected, and then processed in the back office, whereas autonomy obviously is a very real-time application. So there’s some unique challenges there.

But we do a lot of work in the autonomy business. We are the reference system of choice. So when a new autonomy company is forming, they’ll tend to use our system for reference system to train theirs. They’ll use our systems to validate and verify the performance of the autonomous car. And even in some cases, we’re in a production environment with them as well. ⁓ And they’re using our system in a high volume case.

Other than those two big ones, mean, there’s lots of kind of adjacent little markets. We do things like pavement management and road inspection. There’s always a couple of little oddball scenarios here and there, but that’s what keeps it fun. But those are the big segments that we follow.

Stewart Walker (13:11)
Now you mentioned mobile mapping systems and many of our listeners are familiar with the expensive high performance ⁓ mobile mapping systems developed, sold and supported by companies such as Trimble, but also Hexagon, RIEGL and several others. Now we know that Hexagon has its own GNSS IMU solution. Trimble must use your products and maybe Regal does as well. that those high end systems constitute then a significant market for you.

Kevin Andrews (13:44)
They do, definitely. Those systems are absolutely dependent on robust and accurate navigation solutions. And so they are dependent on an accurate system, they’re dependent on a system that works in high volumes. They’re our customer group that really benefits and responds to a high accuracy, robust system. And so, yeah, that’s still a big part of our market. ⁓ It’s still a very much vibrant market.

There’s always new technology is coming online, lidar is getting better, cameras are getting better, the integration of lidar and camera and nurse navigation is getting better. ⁓ But the demands on high accuracy mapping are also increasing. We expect more and more out of our ⁓ commercial map solutions or ⁓ DOTs expect more and more out of their highway management systems. And so they’re putting more and more emphasis on mobile mapping as well. I think the demand for mobile mapping is growing substantially and is still a very relevant business for us.

Stewart Walker (14:44)
At the top end, guess these systems use your POS LV products and our systems. So I’m interested and you’ve already maybe touched on it in terms of interrupted incoming data. But how does the POS LV systems differ from the POS AV systems used on aircraft?

Kevin Andrews (15:06)
Yeah, so fundamentally I would describe them as using the same core components. ⁓ We use the same genus as technology that we get from Trimble or that’s developed across Trimble. A lot of the IMUs and the airborne products and the land products are common, so we use the same things. In airborne systems, especially as the airborne mapping business moves more and more onto drones, there’s a huge push for much, much smaller and lighter weight systems. So, know, small single board solutions, smaller and smaller IMUs.

There’s less of an emphasis on that for land users, so we can use bigger, more robust solutions. When we’re doing something like ⁓ a rail application, putting it on a rail car, they need something that can take more of a beating. ⁓ Same thing in mining autonomy applications. Our customers are looking for something a little bit more robust, ⁓ but the core technology is the same.

I mentioned also that we can take advantage of the road vehicle and some physics, so we can do things like put a real speed sensor on the car or some kind of velocity sensor to get a very precise measurement of the speed of the vehicle, which is easier to do on a vehicle than it on an airplane. So we can take advantage of that. We’ve got more space on the roof. So two antenna solutions are much more common on land applications. These are the fundamental differences. But at its core, it’s the same technology. It’s mostly the same algorithms and the same hardware.

Stewart Walker (16:30)
Okay, now you touched on something there that I was going to ask later, but let’s cover it now. One of the things that I’ve found amazing about Trimble of Applanix, on the airborne side at least, is the development of the APX product range, offering high quality solutions in a much smaller footprint, single board, and a lower price point than the POS systems. And that’s really been a key to phenomenal success in the context of UAV lidar and UAV photogrammetry. But there isn’t really an analog then in the land systems area. just, as you’ve said, you don’t really need smaller, lighter solutions.

Kevin Andrews (17:13)
Well, that’s not completely true. So we do have the same or similar products for the LAN solution. So we do have single-board solutions, we call the APX product line and the LVX, which is a small boxed version of the same thing. So the IMU, the receiver is all on one board and then you can have it in a boxed sealed unit or you can just buy it as a board if you’re a system integrator. We use a dual antenna variant. In Airborne, they have single and dual variants.

But we do see some need for smaller systems. ⁓ The field robotics applications, robots are getting smaller and smaller. There is some drive, not necessarily to reduce the weight, but certainly to reduce the size of it. ⁓ And there’s always some emphasis on cost as well. And so smaller tends to be cheaper. And that’s definitely a hard requirement in some low cost applications. So we do have the same small compact solutions. We just have a more, very neat, I guess, depending on the full portfolio.

Stewart Walker (18:18)
Okay, yep, that makes sense. Thank you. I understand better. Let’s move to the side, so to speak. What is special about solutions for rail vehicles? Or are they the same as you use on mobile mapping systems mounted on road vehicles? I saw ⁓ that you have a product called POS TG for this customer group.

Kevin Andrews (18:44)
Yeah, yeah. So we do use the same POS LV for a lot of rail applications. So for example, you know, talk about Trimble mobile mapping systems, the MX product line, which has Applanix technology inside, does get mounted to rail cars occasionally. It does get mounted onto what the call those high rail trucks, which look like a pickup truck, but have the rail wheels that you can lower so that you can drive them on rail lines. So the POS LV product does go on trains. We do tend to lean more towards some of our IMUs that are built to take a beating. Some of them have higher dynamic ranges. Some of them have higher tolerance to sudden shocks. Because certainly when you’re on a train, when they’re switching rails, it’s not as smooth a ride as it is on a car. So there’s that. And then there’s the POSDG, which is a very specialized product that we developed a number of years ago for Plasser. Plasser being a company that does track geometry and rail inspection work.

And they developed what they call the OGMs, the Optical Gauge Measurement System, which is kind of like a laser profiler that’s very much focused on just the two rails and the relative position of those two rails to each other. They add our GNSS-INS technology onto that. And now you can recreate very, very precisely the relative position of the rails, the curvature of the rails, the absolute curvature, the relative curvature, super elevation, which is sort of the hills and dips, the smoothness of the ride because there’s a lot of focus on rail health, I guess you could say, to make sure that there’s nothing degrading and there’s no possibility of a derailment, which is the biggest fear.

So the POS-CG was developed specifically to integrate with this OGMS and in real time, computes a lot of these curvature calculations that are specific to rail inspection and track geometry and is used for ⁓ measuring the health and safety of the rail environment.

Stewart Walker (20:38)
Yeah, I read an article the other day. It’s remarkable how many countries now have got a train systems that can travel at more than 300 miles per hour. And I’m sure the rail health is quite important in that context.

Kevin Andrews (20:53)
Absolutely.

Stewart Walker (20:56)
So whenever I think of Trimble, one of the memories I have is going to the outdoor demos at the Trimble Dimensions events in Las Vegas. I found it spectacular to see excavators, bulldozers, graders, pile driving machines, and so on, working without human operators. And now I see that one of the products in that area is Trimble RT 200. Are you involved in that in any way?

Kevin Andrews (21:23)
Yes, that is an Applanix product. ⁓ It’s developed here in Richmond Hill. It uses our core technology. ⁓ It’s really just a different packaging. So it was a ⁓ all-in-one box that was developed specifically for field robotics applications. So it’s got a fairly robust IMU in it. It can take a fairly wide dynamic range of inputs as the dual antenna solution and runs all of our technology for integrating into perception systems, autonomy systems and it’s really quite a robust all-in-one package. So yeah, that is an Applanix product. The branding is a little different on it because we were targeting a different market with it.

Stewart Walker (22:03)
Sure, yes. But now also, I looked at your website again before this podcast, I noticed that two application areas that the website talks about are agriculture and pipelines. Now, obviously, these are addressed by airborne surveys to some extent, but there’s also a major land-based component. And I think these areas are of great interest to many of our listeners. So can you say a little bit more about these two segments?

Kevin Andrews (22:30)
Yeah. So the agriculture industry, think is a great example of how these airborne and land technologies really can come together and different applications can really come together and that you kind of need these different applications technologies to come together for the agriculture industry to really take advantage. And what I mean by that is, the airborne side, we’re very interested in applications for detecting plant health as an example using imagery to look at literally at the color of leaves and see if they’re healthy, if they’re getting enough water or not enough water, if it’s harvest season or not.

And so that technology comes from our airborne remote sensing and whether it’s done from drones or crew airplanes. ⁓ That’s great. And so that is the data collection portion of things. And then you move over to, how do we take advantage of that? And more and more farms are looking at autonomy or at least smarter and smarter machines where fertilizer is being distributed automatically based on where you are, based on the information you got from those aerial surveys.

It could be, or the precise farming applications where rows need to be very, very perfectly aligned so that you can maximize the number of rows in your space. And you’re not subject to a driver’s attention being slightly unfocused or being biased in the way they operate the vehicle.

So yeah, so you’ve got, you know, a planet’s working on the airborne side of things to do the data collection and a planet’s working on the autonomy side of things to do the taking advantage of that information. It really works hand in hand. And I’d love to use that as an example of all of these different applications coming together. The pipeline business on the other hand is a little bit different. we’re currently sort of the pipeline business.

We’re in the manner of inspecting corridors, making sure that from an airborne sense, making sure that plants aren’t overgrowing these corridors or that there hasn’t been any landslides or erosions that could threaten anything that’s in the corridor. But many years ago, and this is more of a history lesson, many years ago, Applanix had a problem called the pause pig. Pig being a ⁓ common term for those modules that they actually put in the pipeline and send down the line to inspect the quality of the pipe from the inside out to make sure there’s no cracks, make sure there’s no rust.

And so in the early days of Applanix, we developed inertial navigation systems specifically for this pipeline PIG application. We don’t do it anymore. If the business sort of went away for, and wasn’t our focus over the last couple of years, but that’s one of the places where Applanix got its start in the commercial business was building these pipeline PIG inspection bots, or forward pipeline PIG inspection bots.

Stewart Walker (25:14)
And I apologize to you and to listeners for repeatedly going back and mentioning airborne, but several suppliers have mentioned to me over the last few months, the increasing emphasis on fast deployment and fast data provision. We’ve seen this at INTERGEO and Geo Week as well. This is obviously an advantage in areas such as first response and disaster management.

But, you know, sometimes cynically, I wonder whether we got to a point where fast is always thought to be good, and maybe it’s not. But my question underlying this is, whether there’s anything like that on the land system side, the need to generate data really quickly, even if it’s not at final accuracy. And that enables the user to do some initial work before the final results come in. Is there anything like that on the land side?

Kevin Andrews (26:07)
There are some direct parallels in engineering inspection and that sort of thing where we use a high-end mapping system, and all cameras, lidar, they’ll go through an area, and they’ll very quickly want to know, ⁓ what is the general quality? What do the signs look like? What does the bridge look like? Looking for large areas of damage or making sure that all the signs are accounted for, that sort of thing.

But then when problem areas are detected, they’ll want to go into the more precise information. So they’ll want to get the lighter data out and they’ll want to go in and say, well, okay, well, how big is that crack in the side of the bridge? Or how much is the road rutted and worn through the section? And that requires a higher level of accuracy. So that’s definitely that same kind of like two parts or bilateral sort of accuracy requirement where you’ve got different needs less of a time crunch there.

The first responders in the airborne markets, they’re responding to an incident that’s just happened, like hurricane’s just struck and they need to prioritize the emergency responders. And so there’s a real scenario there where minutes matter. And we don’t really see that on the land side. So I don’t want to make a direct comparison there, but we do see this two-tier workflow kind of come out. The other side of that though is we’ve done a lot of work to make sure that we can respond in minutes or hours for the first responders.

But there’s been some secondary advantages of that, which is on the land side, the data volumes are the problem. ⁓ You can’t fly high with a big wide-angle camera and collect huge things. You’ve got to drive up and down every street. You’ve to drive up and down every alley. And to your point, users expect that information to be updated on a fairly regular basis.

This means that POS-LV systems are running 12 hours a day, every day. ⁓ And if you don’t have an extremely efficient pipeline, that data gets backed up and it never gets out the door. ⁓ So while we’ve developed things to respond really quickly, there’s a secondary advantage, which means that we can also be extremely efficient because we know how to process that data as quickly and as efficiently as possible. And that helps users get through their backlogs and their processing jams.

Stewart Walker (28:26)
Indeed, people often talk about huge data volumes these days, but in the end of the day, you have to solve that problem in some way. And you’ve just explained how. Now let me ask you a question that I guess is on everyone’s lips these days. How has AI affected your software?

Kevin Andrews (28:44)
Yeah, that’s the big question for sure. Trimble as an organization has been extremely dedicated to taking advantage of all of these developments in AI. And so, as part of that, we’re definitely using AI on the backend when it comes to, our developers are using all of the coding tools that are out there today that will help them become more productive and efficient. ⁓ Our customer support team, for example, is using AI to sift through

ask emails and correspondences to make sure that we’re answering questions quickly and efficiently, and in many cases, in an automated way. So there’s of things that we’re doing in the back office around AI. We haven’t quite got to the point yet where we’ve found an application for AI in our product. The GNSS-INS combination is just perfectly suited for common filter, and we haven’t found that there’s anything more efficient than that yet. But there has been some research in

using AI to solve some of these problems, ⁓ especially around the jamming and spoofing issue. Are there opportunities there to come at it from a more creative angle as AI tends to do? We’re using AI very generally here, whether you’re talking about agentic AI or large language models or anything like ⁓ that. Yeah, day-to-day, maybe not so much in the Applanix technology that the customer will see other than the back office and support features.

Where our customers are really seeing AI though that’s more interesting is just a little bit downstream in the data interpretation. So they’ve got a trajectory, they’ve got their sensor data. Okay, now what? It’s just a bunch of points or it’s just a bunch of images. It’s pulling useful knowledge and data out of those that AI is really sort of taking off for our users in terms to interpreting point clouds, turning point clouds into models, looking for things like, find me every sub sign or that sort of thing.

So there’s quite a bit in the industry and it’s really fascinating. It’s a really interesting topic. I’d love to talk a little more.

Stewart Walker (30:45)
Yeah, and I’m sure if we are fortunate enough to be able to talk again in two or three years time, that question will have a much more full answer.

Kevin Andrews (30:55)
Yeah, it’ll probably be your AI agent talking to my AI agent. We’ll just sign off on the conversation after the fair.

Stewart Walker (31:01)
Hopefully they can organize lunch for us,

Kevin Andrews (31:05)
That’d be great.

Stewart Walker (31:06)
I had great conversations not only with Steven Woolven but also with Boris Skopeljak, the vice president geospatial sector at Trimble. And both of them emphasize the increasing importance of products that are offered on a subscription basis. And indeed, in some cases, it’s possible to specify what accuracy you want. Then the software product and the subscription price change accordingly does that thinking affect the land systems too?

Kevin Andrews (31:41)
It does, yes. ⁓ So we do have some products that are available for subscription now. And I will sort of say that there was some pushback from the industry at first. And so for example, POSPAC, which is our post-processing, back office post-processing sort of software, we have slowly been moving away from the perpetual license model and more into a subscription model. There’s some real advantages to that, but we’re hearing from the end users that they want that because they really want to be able to tie usage to their costs, right?

You don’t want to pay for something in a big lump upfront and then try to pay it off or amortize it over years and years. makes the, you know, there’s some accounting advantages to being able to pay for it when you’ve used it and don’t pay for it if you’re not using it. But our business model is pretty complicated. We’ve got a lot of, you know, between us and the end user, there’s dealers, there’s system integrators, there’s a number of different players. And so to be able to get subscription models really distributed the way people want them to, we had to get everybody kind of in step at the same time.

to make sure that our system integrators and our dealers were all able to support subscription models. And so that’s been a bit of a learning for us over the years. So I think we’re over most of the humps now, at point where we can be more focused on making subscription product offerings. Now, Applanix has just one hardware as a service, which I think is really interesting. The idea of only paying for the system as you use it. So that’s on the airborne side. And you love to talk about airborne. I can’t help it myself!

The Applanix PX1 is currently our only hardware as a service hardware product. But that is something that we are definitely exploring on the land side. If there is an interest for that, if it makes it easier for system integrators to scale, to get into different product areas, if it helps distribute some of the risk, we’re certainly getting the feedback that there’s interest in it. We don’t have a product yet that is a hardware product, but all of our software is definitely available in a subscription model at time.

Stewart Walker (33:37)
So I want really to ask you an open-ended question. All of us in Geomatics get really excited about the incredible technology that keeps on incrementally improving. Are there any recent use cases that you’re particularly proud of in the sense you’ve enabled a customer to accomplish something that’s both challenging and important in some way?

Kevin Andrews (34:05)
Yeah. So, you know, I’ve been doing this for a while and we’ve been doing part of the autonomy business since the start. so watching, watching autonomy, just autonomous vehicles go from something that nobody believed possible 20 years ago to, you know, how common it is today. So in general sense, just autonomous vehicles is itself is something that I’m very sort of proud to have been a very small part of. Um, but there was an application that we’ve been talking about very recently that I think is, is absolutely fascinating. And it’s a, there’s a company called Target Arm.

And what these guys do is, they build drone systems and drone landing systems on mobile platforms. So think about landing a drone in the back of your car, which sounds fairly straightforward. And it’s a concept that we’ve seen in science fiction and action movies for decades. swear, I remember the GI Joe’s and the eighties landing stuff on the back of vehicles, but it turns out that’s actually extremely hard to do in practice. And these guys have cracked it and what they’ve done is they’ve learned the hard way that you really have to make sure that everything is speaking the same language. And so they’ve got, and by language in this case, mean datums and I mean timing.

So, you know, for a drone to find a vehicle that’s moving down a road, maybe that road is covered in trees sometimes, maybe that road, maybe there’s even an occasional tunnel in it. So for the drone to find the vehicle is very difficult. If you’ve got any kind of discrepancy in terms of using assessed corrections in terms of timing between the drone and the system on the car. And so the way they’ve solved it is they’ve simply put an Atlantic system on both. And so there’s a PX1 on the drone. And the PX1 is using RTX as its real-time correction service. So the drone is 10 centimeters, 5 centimeters accurate in space, good orientation solution. And then there’s an LVX on the vehicle itself. And the LVX is doing the same thing. It’s also using RTX corrections. It’s in the same data, also measuring its speed, velocity and orientation very precisely.

And that allows the two of them to really come together ⁓ in space. And so these systems get the drone super close and then when the drone is close enough and it’s matching the speed and the motions of the vehicle, an arm literally pops out and grabs the drone out of the sky. ⁓ It’s amazing to watch. It’s kind of uncanny. So it’s something that I get excited about as an autonomy geek and old science fiction fan. To see these kinds of things come to life still excites me. It’s a fun business to be in.

Stewart Walker (36:31)
Yes, I agree. And you’ve mentioned autonomy on a number of occasions. So let’s just before we close, talk a little bit about autonomous cars. I mean, we’re getting closer and closer. I have a neighbor here who just loves the ⁓ autonomous driving and his ⁓ Tesla. I find it a bit scary sitting with him. ⁓ But firms like Waymo are being granted permission to operate in certain cities. Now I can understand that cars designed for the mass market that can be fitted with cameras costing maybe tens of dollars per unit. Lidar costs at the moment maybe hundreds per unit. But what is the role for Trimble Applanix in autonomous cars where the market conditions are forcing price levels down?

Kevin Andrews (37:21)
Yeah, so this is a fascinating problem. ⁓ like I’ve said already, Applanix was heavily invested in autonomy 15, 20 years ago. We have been integral in a lot of the autonomous car development programs over the years, some of the really kind of the biggest ones out there using Planix technology. ⁓ But we don’t really see Applanix products that you know, the Puzzle Vs in those robotaxies and in those high volume applications because like you say, there’s enormous cost pressure to make it work with the sensors they’ve got or work to get those sensors down.

And they’re chasing dollars. They’re chasing every penny they can. ⁓ But where Applanix has been really active is in the training and development. And so that’s where we get in. So we get into an autonomy company very early in their inception. ⁓ And in the early days, a new autonomous car has no idea where it is, no idea how to drive itself, it doesn’t know anything. And so it needs to be trained like all AI. So there needs to be a training program.

And so that’s where Applanix starts is we become the reference system that these autonomous cars learn from. So we’re the precise measurement of position orientation as a human driver drives it around and the autonomous system gets to learn. So that’s where we get our start. And then the next phase is, okay, the cars are getting smarter. They’re starting to drive themselves but there’s a huge validation process. so typically an autonomous vehicle developer will develop a fleet of, I don’t know, let’s say 100 units and they’ll drive them around the city or they’ll drive them around their mine. Maybe it’s a mining autonomy application. And then they’re kind of weaning themselves off of the high accuracy positioning system, accuracy appliance system, and starting to go out on their own. And this is where we’re in more of a record and validate and verify product.

So it’s not quite the high-end high-accuracy systems, but it is kind of our low to mid-range systems that are still accurate and still reliable and still good enough to be a baseline for validating the autonomy systems. But when you get into the high-volume stuff, when you get into the Waymos or the example that Trimble likes to talk about is the Super Cruise in Catalyst. Super Cruise uses a slightly different technology from Trimble, which is ProPoint Go and ProPoint Go is GNSS, INS ⁓ using automotive grade sensors supplied by the automotive manufacturer.

And what I like to talk about is the fact that this ProPoint Go has a lot of its origins in a Applanix’s early days. when Trimble acquired Applanix back in 2003, there was some effort to disseminate the technology that Applanix developed into a number of other Trimble technologies. And so it’s alive and well today in a software solution that some of these autonomous cars are using every day to drive on public roads. So, you know, that’s an interesting part of the Applanix legacy that exists today in those high volume applications.

Stewart Walker (40:17)
Yeah, so it’s not quite Trimble Applanix under the hood, but it’s Trimble Applanix there in the automotive development process.

Kevin Andrews (40:28)
Yes, it’s Trimble under the hood and definitely there’s some of the Applanix DNA in it for sure.

Stewart Walker (40:32)
And I think it’s changing. I read an article last week. I think it’s not sure how to pronounce it. Who, why, you know, the controversial manufacturer of components that we’re not so keen on in the US. They are now offering lidar solutions at very low costs for automotive applications. So that could change things as well.

Kevin Andrews (40:55)
Yeah, the push for automotive autonomy has definitely had an influence on the mapping industry in terms of driving down sensor costs and making sensors smaller and cheaper. And it’s had a fascinating influence on the business, which you can do for a lot less money than you could 10, 15 years ago.

Stewart Walker (41:14)
yes, we’re in an exciting place at an exciting time and maybe that’s a good point to draw things together. So what can you tell us about Trimble of Planets, these plans for the remainder of 2026 and indeed a little bit beyond that, what can we expect on the land system side?

Kevin Andrews (41:34)
A lot of our focus and recent product releases have been moving the Applanix, the POS LV technology onto a newer hardware platform, the LVX Plus, which is all the same technology, but smaller, lighter, cheaper. So there’s going to be a big push to make sure that rollout happens smoothly and that customers could take advantage of that. Internally, we’re very focused on some of those workflows.

We talked about how sensors from Autonomy are making their way into mapping, but some of the smarts behind it as well. And so we’re coming up with products like Lidar QC, where we’re using SLAM technology and we’re using perception technologies to improve the mapping workflow so that we can do things like calibrate your cameras and borset your cameras, calibrate your lidar and boreset your lidar in a much more streamlined and efficient way. ⁓ We’re really kind of working on that ease of use problem for these big integrated systems. And that’s something that we’re putting a lot of emphasis on and we’ll continue to do so in the near future.

Stewart Walker (42:35)
Okay, thank you. So Kevin, Andrews, I’ve enjoyed our conversation. I’m so grateful that you were able to participate in the Lidar Magazine podcast series. It’s been really interesting to talk about the land system side of Trimble Applanix And I hope that to some extent we’ve redressed the balance, which was rather too heavily ⁓ on the airborne side. Readers interested in GNSS IMU capabilities for their land-based applications are now much better informed. We wish you well with growing this business and the technology in the years to come and to ⁓ hearing more about it in this podcast series and writing more about it in the magazine.

Kevin Andrews (43:19)
Thank you for having me. has been great. And it’s measuring the world. It’s a crazy puzzle and I love it. So thanks for letting me talk about it for a little bit.

Stewart Walker (43:28)
Well, I’m sure listeners will similarly have enjoyed your company and comments. I want to underline our gratitude to our sponsor, the popular LAStools lidar processing software. And we hope that listeners will join us for forthcoming podcasts. We’re expecting some guests whom we believe you’ll want to hear. If you want to ask about our podcasts or make comments, don’t hesitate to write to podcasts at lidarmag.com.

{Music}

Announcer: Thanks for tuning in. Be sure to visit lidarmag.com to arrange automated notification of new podcast episodes, subscribe to newsletters or print publication and more. If you have a suggestion for a future episode, drop us a line, here. Thanks again for listening.

{Music}

This edition of the LIDAR magazine podcast is brought to you by rapidlasso, our flagship product, the LAStools software suite, is a collection of highly efficient, multi-core command line tools to classify, tile, convert, filter, restore, triangulate, contour, clip and polygonize lidar data. Visit rapidlasso.de for details.

THE END