Subscribe
Share
Share
Embed
Demystifying the conversations we're already here at RRE and with our portfolio companies. In each episode, your hosts, Will Porteous, Raju Rishi, and Jason Black will dive deeply into topics that are shaping the future, from satellite technology to digital health, to venture investing, and much more.
Raju: Jason remembers everything. He is a robot, by the way. People don’t know this.
Will: Welcome to RRE POV—
Raju: —a show in which we record the conversations we’re already having amongst ourselves—
Jason: —our entrepreneurs, and industry leaders for you to listen in on.
Jason: Matt Rendall, CEO of Clearpath Robotics & OTTO Motors, their industrial subdivision, welcome to the podcast. Really excited to have you here, especially after the exciting year you guys had last year. And we’re excited to dive into how you built one of the most successful robotics companies in the last decade. So, we’d love to kick things off with just, like, a brief background on yourself and how the, kind of, initial company came to be before we dive into the evolution.
Matt: Yeah, thanks for having me, Jason. So, I studied Mechatronics Engineering at the University of Waterloo, and at the time, it was a relatively new discipline. I found myself a little bit bored on an internship, and to fill the void of, kind of, mental stimulation that I was seeking with the internship that I wasn’t getting, made my way to the robotics lab and the robotics club. This was kind of early on in my undergraduate experience, and just immediately was addicted to mobile robotics, and this was, kind of, around the time that the DARPA Grand Challenge was beginning to capture the hearts and minds of engineers and scientists, and so yeah, just started building mobile robots, competing in international robotics competitions, primarily in the US, but sometimes a little bit further abroad, and we even won handful. And so, when it came time to graduate, the goal was, let’s just continue building robots.
And at the time, you couldn’t get a job in mobile robotics. It just wasn’t really an industry, yet. So, I sometimes reflect back on the decision point to start a company. I’m not sure if it was more forced entrepreneurship than an active decision. You know, we just wanted to build robots, and we couldn’t get jobs doing it, so we had to create jobs doing it. And that’s, kind of, how we got started.
Jason: And so Matt, just for the benefit of our listeners, before we get too deep into the story, why don’t you give a quick overview of what Clearpath Robotics does, as well as what OTTO Motors is.
Matt: Clearpath and OTTO Motors are autonomous technology companies, both operating under one legal entity. So, Clearpath Robotics focuses on autonomous technology and platforms for research and development, helping innovators speed time to market, reduce development costs, and mitigate technical risk with the development of novel autonomous mobile robots. And OTTO Motors is our industrial division, taking all of that capability and verticalizing it into an autonomous material handling offering for the world’s largest manufacturers, saving them time, improving safety, improving costs, and just creating a generally more resilient indoor supply chain for their manufacturing operations with autonomous vehicles.
Jason: And just for people [laugh], just for the benefit of the people, since it’s an audio medium, they’re like, just little robots, like, that are kind of like, have a flat platform [laugh], and four wheels and some sensors, and you can, like, build on top of them, and have different effectors and actuators, et cetera. They’re like a baseline platform for navigating around the world, whether that be outdoor and indoor, for Clearpath, which are little yellow robots. I shouldn’t say little; some of them are kind of big. And same thing, like, OTTO Motors is much more sleek, much more, kind of, industrial-hardened, but still the same thing: four wheels, navigating around.
Matt: In the OTTO Motors product portfolio, our smallest robot is about the size of a cart. It’s designed to move carts. So, if you’ve been in—for anybody—
Jason: Yeah, it’s an ottoman.
Matt: Who has been in—
Jason: It’s like the size of an ottoman for a couch.
Matt: Yeah, basically, it’s an ottoman, though, it’s tasked with pushing carts around the factory. So, that replaces, kind of, human-scale manual pushing delivery routes. We’ve got a robot that’s about the size of a pallet, which moves pallets. We’ve got an autonomous fork truck, which is the size of a fork truck. And then on the Clearpath Robotics side of things, we’ve got robots that range from small, approximately the same size, like, one the size of a laptop, all the way up to something the size of an ATV or utility vehicle. And then occasionally we get into building more specialized, larger-scale autonomous vehicles for our key customers.
Jason: And truly a platform. It’s a platform for other things [laugh].
Matt: Yeah, exactly.
Jason: Both hardware and software. Yeah.
Matt: Yeah. Our core is the industrial autonomy so that the control platform to allow an industrial machine to intelligently and safely navigate its own way through an environment to complete whatever that mission if it needs to complete is, whether it’s moving a pallet, or moving a sensor payload for substation monitoring, we have the autonomous mobility platform to do that job.
Jason: Yeah. Give us a little bit more of a sense of, like, the state of play in the robotics space, at the time, when you were starting out. I mean, the DARPA, those of you who know, like, that’s really led to a lot of autonomous driving, and what would eventually become Waymo, et cetera. Like, you know, set the scene a little bit for the state of robotics when you founded the company. And yeah, we’d love to hear a little bit more about that.
Matt: Yeah, well, you know, the DARPA Grand Challenge, I think, is probably a good thermometer. At the time, the first DARPA Grand Challenge—I can’t reme—it’s been quite some time since I’ve looked at the actual profile, but you know, it was you had to drive something like 100 miles across the Mojave Desert. They chose the Mojave Desert because it was, you know, harsh, unforgiving terrain, but it didn’t have a lot of the complexities associated with urban driving. And in the first competition, the furthest any team got—and we’re talking about the smartest roboticists from all the top Ivy League schools were competing in this—the top school got, I think, three miles or four miles, and then failed out. Everybody else failed out sooner than that, right? So, that was the state of autonomous driving around the time that we got our company started.
Raju: Did they, like, take a brick around the gas powered car and just wrap it around and see where it went?
Matt: Yeah. You know, you might have actually gotten further, if you had just taped a brick to an accelerator. You know, if I reflect back on the time, really one of the founding value propositions that we saw an opportunity then was, roboticists would spend 80% of their time trying to configure up the hardware, only to afford themselves 20% of the time to actually develop the algorithms which were the state of the art, and what we wanted to do was allow for an inversion of that. We would build the platforms to allow for 20% time spent to configure up the hardware, and 80% of a PhD thesis to be spent on progressing the state-of-the-art in autonomous control or autonomous software development.
And at the time, open-source was really only cutting its teeth in autonomous vehicles as well, and we were the first company to support that. We thought, you know, if we really are going to foster the state of the art and accelerate the state of the art, open-source is going to be a critical ingredient in that. So, we plugged into that very quickly as well. And yeah, very quickly, very successful with that thesis. We became kind of a market leader in this really nascent, early component of autonomous technology development, by offering, essentially, the innovators in the market the ability to spend more of their time innovating.
Jason: That was actually where we ended up bumping into Clearpath was in the visiting campuses and robotics centers at MIT and Carnegie Mellon, et cetera, we kept seeing these yellow robots, these Clearpath robots [laugh]. It’s like, “Hey, maybe we should go talk to those guys.”
Matt: Yeah.
Jason: And so, I know that both on just, kind of, the awareness side, but I thought this was also an amazing—maybe we’re skipping a little bit forward, but you had an amazing story around the talent pool as well, that also drove.
Matt: Yeah. Well, you know, I think RRE and Clearpath ended up colliding because we both identified the same, kind of, first principle, maybe we can call it, that a really deep technology will incubate in research and development for a protracted period of time, and so you’re going to start seeing those early market signals in universities, in research labs, and it’ll exist there for a much larger period of time than maybe some less deep tech domain. And that was—you know, when RRE started to identify robotics and mobile robotics or autonomous vehicles as an interesting category, to your credit, you were looking at that theme way before any other VC in my experience was, and you started knocking on the doors of research labs and universities, and that’s where you saw us. A byproduct of that strategy of serving the research market we stumbled into, and it’s still to this day, I think, one of our superpowers, is the world’s talent pool that is proficient in autonomous vehicle development, very high probability that they learned on a Clearpath robot platform anywhere in the world, right? So, when they graduate, and they’re doing their first engineering or computer science job out of school, we’re the brand that they know, and so when the big technology players in the space are recruiting talent, and they’re seeking to get job applications from the top talent graduating, that same talent also will tend to apply to work with us because they’re so familiar with us, and they’ve grown to know our company and our engineers and software developers in the time since. It really allows us to punch so far above our weight class in the talent market, and it’s a very, very precious talent market.
Jason: It’s reminiscent of, like, the Photoshop in design. You know, before it was SaaS, it was all license-based, and they could crack down on, like, design students pirating their Photoshop software, but they’d rather have them all get used to Photoshop and then go into the design world and then say, “Well, actually, I mostly just know Photoshop, so I guess we’ll use that here [laugh].”
Matt: Yeah, exactly right.
Jason: And it’s been huge for both Photoshop and Clearpath. Maybe you can talk about, like, the evolution of the business from, you know, almost a passion project that happened to become a business to, you know, obviously, the development of OTTO Motors, and a much more industrial, hardened, like, scoped use case. You know, how deliberate was that transition? And maybe you can talk us through that path a little bit.
Matt: The Clearpath Robotics story, you know, we were bootstrapped for a very, very long time, an atypical amount of time compared to the startup market, so long that you could possibly take a cynical view that we weren’t actually a startup before we took venture funding. So, you know, if we rewind back to 2008 when we were getting our feet under us, the economy was not in a great place, and the iPhone had just been released the year prior, so the limited venture funding that was available was, generally speaking, much more interested in mobile app development. And, you know, wouldn’t touch with a ten-foot pole mobile robot development. By necessity, we had to figure out ways to get customers to fund our business plan, and you know, we did do a very small round of angel funding about 12 months into our pursuits. We ended up getting a handful of purchase orders, and we couldn’t finance those purchase orders.
Banks wouldn’t touch us because we didn’t have a track record or any—you know, as recently graduated students with zero net worth, we weren’t terribly creditworthy. So, some angel investors, our seed round was essentially PO financing, and we, over the course of 12 months, executed those purchase orders, became profitable, and grew our business for the next six or seven years, aggressively and profitably. And it was during those six or seven years that we developed, not only credibility, but really potent insights about the market and where it was going, and that allowed us to—serendipitously, right around the time that we were starting to think about what our next big focus would be—collide with RRE and begin the venture-funded journey. We started to see signals that the technology was leaving the research lab and entering industry. We saw an opportunity to be a bigger part of that transition, and so our aspiration at that point was to move from being a provider of technology to the innovation market to being a provider of autonomous solutions to industry. And in order to make that leap, we had to become venture-funded, and then that, you know, that’s when we got connected with RRE. You were our first institutional funding, our first venture funding round—
Jason: Almost exactly a decade ago.
Matt: In 2015, yeah. And then, you know, the rest is, as they say, history.
Jason: Oh, I guess-yeah, nine years ago, sorry. Talked to you guys in late-2014. Round was done in 2015. But boy, what a journey from there. I wrote up a little piece for the Rockwell announcement, recalling what a deliberate, kind of, hierarchy of needs that you had laid out during our, you know, diligence conversations. That was such a [prescient 00:13:35] way to think about the company. I’d love for you to talk a little bit about how you were thinking about it at the time, before OTTO Motors really was a real thing, and how it helped guide the path forward for that division.
Matt: Yeah. Actually, and I saw the article. Your memory was really good. You had it pretty darn close from memory. Yeah, so I mean, the thinking was that, over time as the market develops, the most valuable and durable thing that we could offer the market would be in the intellectual property domain. And that was, you know, one step above software, which was one step above the supply of systems, which was one step above—at the base of our pyramid—the widgets or the components that you could configure into various different test systems for agricultural research and development, or mining research and development, et cetera. And so, we just crafted a strategy which would allow us to climb up that pyramid as the market matured. We sought to play in the layer that was most durable, most valuable, most profitable, you know, I think even—we’re 15 years in, we’re 10 years into our relationship with RRE—even 15 years in, the market is still very early.
Jason: Yeah, I was about to ask, do you think that’s changed for a new robotics company coming in? Like, how much—you know, you had to build your own hardware, right, is the, kind of, net-net. Like, there wasn’t another AMR platform to build on. In many instances in industrial environments, that’s still kind of the case, unless you’re talking about, you know, robot arm that’s maybe a collaborative robot with some sort of end effector. And you need to do some path planning and picking, et cetera, et cetera. But for the most part, like, do you think, if you had to start OTTO Motors today, that it would be dramatically different in terms of the capabilities that you need to build out?
Matt: A robotics system is still fundamentally a system, and we approach it with a systems view. I think the first thing that jumps out at me in terms of why it would be easier today to be an AMR company starting from scratch, the supply chain offers more out-of-the-box solutions for AMRs. There are, you know, off-the-shelf lithium today, not only is it a thing—it wasn’t a thing for us back when we got started—but there are [packs 00:15:52] that are designed and optimized from an energy density, charge profile, price point, form factor, et cetera, purpose-built for AMR. There are power trains purpose-built for AMR. There are computer formats—form factors—purpose-built for AMR.
So, the barriers to integrating a system, as well as how far open-source has gone, make it so that you could get a credible prototype, you know, up and running far faster for far less dollars. But in terms of, you know, the market that we serve in high-cost-of-failure, industrial environments, the challenge is much less in that first prototype and much more in how do you get to the five nines reliability with the global service and support coverage necessary to deliver the offering to that customer? And that’s something that is hard to throw money at to accelerate, and just really needs to be done thoughtfully, and you need to get your nose bloodied a whole bunch of times in order to figure out where the deficiencies are in order to fix it. And there really haven’t been a lot of—maybe this is an opportunity for entrepreneurs listening—if you can innovate in the service and support side of things, you could offer a lot of value to AMR companies.
Raju: I’ve been in this space a long time. In fact, my son is graduating shortly with his master’s in robotics at Carnegie Mellon. So, he was—he worked for you guys [laugh]—
Matt: [crosstalk 00:17:18].
Raju: —for a summer. Loved it. Really great experience, actually. And I pick his brain a little bit, just on different flavors of, is it getting easier? Is the rate of innovation changing? So, early days of robotics, it felt really kind of like a slow-moving path. Even though there was a lot of interest in it, it felt like a slow-moving path, and you know, there was sort of step functions of innovation, as opposed to sort of this geometric or curve-based, you know, process. Do you think the rate of innovation has gotten faster in this phase, do you think it’s kind of the same, or do you think it’s actually slowed down a bit, you know, in terms of the number of people that are focused on, sort of, creating that component-level capability?
Matt: I would maybe look at answering that question from a couple of different vantage points. The first from an enabling technology perspective. There are still a lot of benefits to Moore’s law coming online, in particular in the domain of sensors. So, if you look at what’s possible with sensors today compared to what was possible with sensors when we got started, it’s really, really an exciting curve that’s been climbed. And what that does, when you look at all the different pieces of a robot, and where and how things like Moore’s law have been benefiting it, the form factors and cost points and performance, or capabilities of that robot open up new applications that even five years ago might not have been possible.
I think the battleground for innovation going forward in the next, call it, decade is going to be on the product-market fit, figuring out which applications, where can a very specific robot be built to solve a very specific problem in a fast and impactful way. We are seeing—or I am seeing—entrepreneurs and companies start to try and solve very specific niche problems, and those are the ones that I’m most interested in because those are the ones that I think are going to win quickly, build a foundation, and then be able to expand into adjacent market opportunities. There’s a lot more attention being paid now to mobile robotics by the investor community, which is a critical ingredient, in small part, I’d like to think, because of the success of Clearpath, and showing that you can develop a very healthy return for your LPs by making smart investments in mobile robotics. But there’s a lot of fast follow happening right now in AMR categories that have been proven on that product-market fit innovation domain, and there are so many AMR players in material handling right now, it’s painful.
Jason: Yeah [laugh].
Matt: And I don’t envy what they’re trying to do right now because we have the benefit of ten years, among other things, hardening a global service and support strategy, which in our market, is more valuable to our customers than the technology itself. So, I wish that these entrepreneurs would focus more of their energy towards solving new problems in new markets. That’s where I think there’s a ton of opportunity. But maybe, to your question about, is innovation slowing down, maybe I wouldn’t say it’s slowing down. Stagnating, maybe, as the market works to kind of capitalize on the opportunities that have been validated.
Raju: Well, you’ve said something super interesting because the way I kind of think about markets is, it starts as a point solution, kind of, entry point, and eventually you get to platforms. You know, the way you had coined it just earlier was, I’m excited about thinking about new point solutions, new sort of verticals or incision points. Is this industry still, you know, sort of that level of capability? We’re not really ready for platforms; we’re really, you know, ready for more verticalized, or you know, precise use cases. You know, that’s the first question I have.
And then, you know, just related to that, you know, how bespoke do you need to be in this industry? Because it still feels relatively bespoke, as opposed to, you know, sort of giving people the tools like, you know, if you look at salesforce.com, they don’t really care what industry you’re in. They give you a tool to manage your sales team, and they just say, “Go to town,” you know? “Figure it out. Do your own sort of fine-tuning.” You know, just curious on your thoughts around those two.
Matt: In order to serve an application properly, at least in the use cases and with robots that I’ve been exposed to, it does need to be a very targeted solution. But the good thing is that the market opportunities for those incision points are still massive. And you know, your average sell price for a single robot to a single customer for one of those use cases, is several orders of magnitude larger what it would be in a more commoditized software category.
I’ll use an example that’s, kind of, totally adjacent to what we ended up pursuing. But in the early days of, kind of, market exploration with Clearpath, we started to get really interested in substation monitoring. We had a customer in Florida that was interested in using mobile robots for substation monitoring, and the whole premise here was use a robot to monitor a substation on a daily basis, as opposed to the very high cost and low frequency preventative maintenance surveys that currently happen on substations. And if you can measure a couple of very specific and repeatable things, you’ll be able to better predict failure trajectories on your transformers at your substations, and prevent big outages for your population and expensive reactive repair. That one utility in Florida alone had 11,000 substations, right?
So, that’s one random example that we had come across over the years, where if you wanted to build a robotics company focused solely on being the best in the world at persistent, preventative maintenance monitoring for substations, your addressable market in the United States would be attractive and sufficient for quite some time before you would need to diversify into other high-value asset monitoring. But you could very well go adjacent into high-value asset monitoring from that. I just wouldn’t recommend it at the onset.
Will: Powerful example, Matt. I think it speaks to the idea that point solutions matter a tremendous amount in the current state of the market. It would be great to hear you comment, sort of, generally on the journey that you saw customers go through to onboard your platforms. I mean, I think from a removed perspective of an investor just watching Clearpath’s growth over time, I was under the impression that you had customers engage in very lengthy pilot processes, and that the journey into the production environment took a long time. It would be interesting to hear, kind of, the thought process that you saw customers go through. Because you were laser-focused on addressing their needs in specific point use cases.
Matt: I guess I should preface by saying we expected that it would take a long time, but we undershot how long [laugh]. The manufacturing market was very attractive to us for a number of reasons. It’s critical for our economy, for many economies globally, heavily dependent on a labor pool which is becoming increasingly scarce, so we saw a forcing function coming, right? So, if you could survive the pilot purgatory period that you’re alluding to, on the other side of that forcing function would create a wave of automation adoption that would be incredibly compelling to be at the front of. That pilot process, I think, is an artifact of, yeah, the cost of failure is very, very high, and in industries where the cost of failure is very, very high, adoption of innovative new technologies, solutions, vendors even, is a very slow process because the stakes are just that high.
And in our case, we were a new vendor offering a new product in a new technology category. Everything was unproven. The industry, while manufacturing is very large, it’s also very small in some ways. So, the person making the decision to bring in a new technology from, you know, a new company with a new product and a new technology needs to be very, very certain before they put their name behind it because you could create a lot of headwind for yourself professionally if you make a wrong decision and take down a plant. Though, now that we’re on the other side of it and having the benefit of being on the other side of it, I can look back and say, “Yeah, it makes sense,” it was excruciating as we went through it because where we were at at the time, was still working on that product market validation hypothesis internally, and there was a lot of internal discussion and churn.
“Oh, you know, shoot, did we… did we get this wrong? Are we not looking at this properly? Should we be pivoting to that substation monitoring application?”
Will: Right.
Matt: Because these pilots are just going on and on and on, and you know, when are we going to get the big order from the plant, and when are we going to get ten plants from the account? Finally, it did come—
Jason: Sometimes the numbers are just tempting, right, where you’re like, “We delivered 10x ROI. Like, what else do you want? Roll it out to all of your entire manufacturing floor or warehouse.”
Will: I imagine that’s the case. If I could just ask one more question, what were the qualities of your favorite customers [unintelligible 00:26:52], Matt? Because some of them really did lean in a major way.
Matt: The early-adopter persona for us in our industry, the companies that—and individuals—that really truly understood that an investment in new technology was being made to create a competitive advantage for that business, and doing so had some risks. And when—when, not if—we encountered problems, we would tackle them together as partners because on the other side of those problems, once they’ve been resolved—once they’ve been resolved, not if they had been resolved—there is a competitive advantage worth fighting for. Those are some of the best companies that we had. And, you know, we are where we are because of the, you know, the technology sponsors and the early adopters that managed to navigate through some really risk-averse waters in order to bring us online.
Raju: I have a question for you. Think like an investor for a few minutes—so that means you probably have to drop your IQ by about 30 to 50 points; so [laugh] you just have to pretend like you’re 12—you know, what type of robotics company or two, would you invest in? Like, if you had the opportunity to put money into something now?
Jason: Are you looking for specific names or just category?
Raju: No. No names. No names.
Matt: Just categories? Yeah. Categories.
Raju: Just the type of company. Just the type of company.
Matt: I think I would double down on my earlier statement about a company that’s really focused on a point solution, right? A point solution that may be so specific that most others are not interested in it or not focusing on it, but spend the time to unpack, okay, let’s assume that you’re successful there. What does that earn you the right to do as an adjacent market or an expansion point, and so on and so forth? Those very, very specific applications, I think, should not be underestimated. And then the second dimension that I would add on top of that would probably be looking at how painful is it to get a person to do that task?
You know, one of our earliest mission statements would be—or was—automating the world’s dullest, deadliest, and dirtiest jobs with autonomous mobile robots. What we think is the most compelling opportunity and application domain for mobile robots is doing the tasks that people can’t, won’t or shouldn’t be doing in the first place, right? Substation monitoring, to go back to that example, if you’ve got 11,000 of these facilities spread all across the state of Florida, how many service technicians could you reasonably have doing that work, right? You just can’t get the coverage that you need, and the cost to get the coverage that you need is [off side 00:29:34], so you end up with really low resolution information in order to operate your distribution network. So, making sure that you’ve got a very specific point solution and a very compelling—what’s the right way to—not labor arbitrage, but if the labor component in the equation is very problematic, that’s a good indicator that if you nail the point solution, and you validate the ROI case, people will make the change because there are very often much better, more valuable things for the people that are in that workforce to do.
Raju: Yeah. I know you love your new home at Rockwell, but how can I persuade you to build a personal bartending robot for me?
Matt: [laugh].
Raju: I got a bunch of whiskey bottles that need pouring.
Matt: You know, its—
Jason: You got to support the [PR2 00:30:22], right?
Raju: Yeah.
Matt: Yeah. Yeah, yeah. I think that, you know, there have been a handful of bar-bots that have been made over the years. It’s not uncommon to see that at trade shows. It’s one of the interesting party tricks to pour beer.
Raju: Yeah. No, I needed to actually come to my house, take stuff out of the liquor cabinet because my wife and children refuse to do this anymore. I invite Jason over every once in a while. He pours me a couple, but he doesn’t realize he’s getting roped into, like, labor pool. Uh, anyway.
Matt: Well so, I mean, I know that you’re partially joking here, but that’s a great example of you know, there’s a pursuit right now—this is just my humble personal bias perspective, right—the pursuit to create humanoid robots is not interesting to me. It’s not part of the recommendation that I just gave. It’s our equivalent of artificial general intelligence. It’s not the point solution. If I were to use your use case example with my previous response, I wouldn’t build the artificial general intelligence humanoid robot that could make—you know, walk its way over to your cabinet and then mix whatever cocktail you wanted; it would be something that looked more like a Keurig machine. Maybe there’s a robot arm that could mix four cocktails, not infinite cocktails, right? So, it’s the purpose-built machine for a very point solution is where I think the next cohort of winners will be.
Jason: I’m curious, like, there were a couple of times, just getting back to, kind of, the OTTO Motors journey, you know, we really pushed the envelope in terms of, like, the capabilities, revved the engine. I’m thinking of—I’m hoping you get this reference—we had a tractor company that we were helping for a little bit with a pretty innovative solution. How did you guys find the right balance between pushing an innovative product and finding where the barrier was that pushed beyond where a reasonable person would, you know, hit the buy button, or sign the contract? And I’d love for you to describe the deployment, even if we can’t say the name.
Matt: Yeah, it probably is worth talking a little bit about our Series A plan because that leads into that question. So, where we were when we raised our Series A, we had received a handful of contracts for prototype development of autonomous pallet movers for the logistics market, and this was part of the work that we did core to serving the innovator space, but very quickly, a number of projects came out of left field. You know, we weren’t paying attention to this market opportunity; we were much more focused on harsh outdoor applications. And in fact, at the time, we were doing a lot of open-pit mining projects. Like, the last autonomous vehicle we built before pivoting into OTTO Motors was a 40-ton haul truck doing open-pit mining in Brazil. Like, that, that was, kind of, the type of autonomous vehicle that we would build at the time.
But we got these market signals very strong, very quickly. We analyzed it, and against mining and a couple of others, we said nope, manufacturing is going to be the best. Okay, we need to raise some venture capital, get this done. We had poured every last available dollar that we had on our P&L and our R&D budget to producing the first autonomous pallet mover. It was held together by duct tape and bubble gum on the inside, but it became the demonstrator that, I think, played a big role in helping convince RRE that there was an investible opportunity with Clearpath. And we raised—it was ten million or so, approximately. I can’t remember the exact amount.
The goal was to take that duct tape and bubble gum prototype into alpha stage. We wanted to build nine vehicles, we wanted a couple of them for engineering test and development, and then we wanted to do two or three pilots of two or three vehicles each, so that we could start to validate the operating environment, the business case, which applications would be interesting, you know, very, very early stage alpha type stuff; there was still a lot that needed to be built. So, in order to get these pilots, we had to do some demand generation. We started doing some webinars, ended up getting a much stronger response from the webinars than we were expecting. It very quickly matured into, we had a prospect for a pilot, we wanted to get two vehicles inside of that company, and essentially they said, “Forget your pilot. We want a production system. We’re ready to go. Here’s an order for 16 vehicles, but it’s game day. Like, you’re going on to a production line.”
We couldn’t say no to that. There are many reasons why we should have said no to that, but we couldn’t, you know? We just, we had to take it. We knew it was going to be painful, but it was a really important part of our learning journey. While we were in the heat of that battle, that same pipeline from those early demand generation webinars that we were doing led to a second customer saying, “Forget your pilot. Here’s a production order for 49 vehicles. But it’s production. It’s game day. We got to go.”
Again, we had to say yes, you know? And to frame this, we were expecting we’d be able to get a couple of 100 or $200,000 pilots, and we went to—the first order that we weren’t expecting was like a $2 million order, and then the second order we weren’t expecting was several times more than that. And we got a lot of bumps and bruises. I mean, we were, in many ways, building the parachute after we had already jumped out of the airplane on those two projects. And because we were in a high cost of failure environment, we had incredibly high pressure to keep the robots running.
And in order to keep the robots running, we had to put a lot of people on the robots. We had to babysit the robots, we had people living at the customer sites, we had multiple shifts, we had on-call, we had 24/7 just—
Jason: The pinnacle of efficiency.
Matt: Yeah. Just, we were brute-force keeping these installations alive. And the one that you’re referring to was a really early lesson in, okay, even within a specific application—a specific point application—there are certain use cases that are good for our robots and certain applications that we will never do again, and still, to this day, we have not done again. And so, that was one really critical project in our history. It showed us, more than anything, that service and support was more valuable, and we adopted a mantra at that point that our customers will buy B-grade technology with A-grade service and support over A-grade technology with B-grade service and support. And at that point, we started to realize that we needed to make big investments in our service and support capabilities, and that’s what we started doing around the time of our Series B. I’m not sure if that’s talking about what you were hoping. Happy to double-click on [crosstalk 00:37:17]—
Jason: Yeah. I mean, I was more referring to the magic carpet, which was like, moving—completely replacing this idea of, you know, an assembly line and the products moving dynamically from station to station. Like that’s, like, really pushing the envelope. In terms of materials flow, this was, like, at a hundred percent optimized, and only capable because OTTO Motors owned a swarm of robots that were all controlled by a centralized system, and then giving you unbelievably detailed feedback about how long the product stayed at each station, and how you could further optimize.
Which I think there’s this interesting thing where—you know, we’re going through the same thing with LLMs, where everybody just takes things to the logical conclusion that, like, the future world is the hundred percent version of what you would imagine, but the reality, at least in this instance, is something that pushes beyond what was capable before, but you know, as you kind of said—maybe this is the A and B thing—but, kind of, is 80% of the future, and the rest of the 20% is, like, whether it’s comfort or trust or security or the support that, kind of, like fills in some of that gap, whereas, as a pure technologist and a pure futurist, you might want the magic carpet replacing, you know, all potential assembly lines, and maybe that actually is the most efficient thing, but not only do you have to have, like, unbelievable uptime, and SLAs, et cetera, but you also have to have somebody that’s, like, willing to take a bet and willing to almost risk their career, if you’re going to roll it out, like, nationwide for some of these large manufacturers to actually, like, push to the furthest extent. Which is a little bit antithetical to the thinking of, you know, at least the existing group of people running large industrial environments. For good reason, right?
Raju: Hey, Jason. Do you have a set of Gatling gun questions? I’ve prepared a few if you haven’t.
Jason: Let’s hit it.
Raju: We always end these podcasts with Gatling gun questions, which is really just like single sentence, single word answers. And so, I’m going to start with, best enforcement robot: Terminator or Robocop?
Matt: I would say—I would have to say Robocop from an enforcement perspective.
Raju: Fantastic. All right, fantastic. And the cutest robot: R2D2, Wall-E, or Big Hero 6.
Matt: Oh, Big Hero 6.
Raju: Fantastic. Okay, give me a year—this is a guess, you know? I’m not going to hold you to it—
Jason: I’m going to hold you to it, though.
Raju: Yeah, [look 00:39:46], Jason holds—Jason remembers everything. This guy has a brain.
Jason: Everything.
Raju: He is a robot, by the way. People don’t know this. But when will autonomous transport vehicles be sanctioned on roads in the US? Just a year.
Matt: I would probably say, this is, like, a 2030 thing.
Raju: Okay. How about drones for the skies? Or transport vehicles in the sky? Like, literally transport vehicles in the sky?
Jason: Like VTOL.
Raju: Yeah, VTOL.
Matt: Like passenger craft?
Jason: Yeah.
Matt: Probably a similar timeframe. I’d say 2030.
Raju: When will we have a humanoid robot in the workplace?
Matt: A humanoid robot in the workplace? Well, it’s starting, actually. Well, so in the industrial workplace?
Raju: Yeah. Industrial.
Matt: Within the next couple of years.
Raju: All right.
Matt: You know, so yeah, call it 2025, 2026.
Raju: How about in the home? To basically bartend for me, essentially, but maybe other purposes.
Jason: It’s going to do elderly care, and then for Raju—
Raju: [laugh].
Jason: —he’s going to hack it. So, it can—
Raju: Actually, you know what? You could probably put those to the same bucket. [crosstalk 00:40:54] [laugh].
Jason: Great, because, like, the elderly person needs—they need medicine poured into a little cup.
Raju: Yeah.
Jason: That’s perfect for you, Raju.
Raju: I know.
Jason: You just swap the bottles out.
Matt: Yeah.
Raju: No, no, no. I could actually use it for my medicines, too. I’m actually in the elderly category, now.
Jason: Oh, fantastic.
Matt: So, like the Jetsons, that’s going to take quite a bit more time on Moore’s law. I’d put that more like 2040, 2045. Like I say, it’s going to be some time out before it’s practical.
Raju: Okay. Well, that’s the only ones I had. I don’t know if you had any Gatling gun questions or others, Jason?
Jason: Oh, no. You’re Mr. Gatling.
Raju: Oh, I’m the Gatling gun?
Jason: And you surprise our guests, too. It’s just like—
Raju: I do.
Jason: Completely… completely out of nowhere. By the way, get ready.
Raju: [laugh].
Jason: [crosstalk 00:41:35] Robocop.
Raju: Robocop.
Jason: But anyways. No, I mean, I know we’re coming up on time. We’re a little over time, actually. But, Matt, I mean, it’s been an absolute blast, you know, being alongside you for this journey. I’ve learned so much just being a part of the story, and you know, through the ups and downs. It’s been a real pleasure to be to play the part I have and for RRE to be along and support you, for truly one of the most successful robotics companies of the decade, which is a huge achievement, and kudos to you and the whole team for sticking to it and really getting through the hairy mess [laugh] of an industrial sale, and the materials transport. So, bravo.
Matt: Yeah. So, I’m super grateful for the partnership that we had, and I think still have, with RRE. I mean, mobile robotics was a very unproven—not only an unproven market opportunity from an entrepreneur perspective, but almost an even riskier bet from an investor perspective, given everything that had to be done subsequent to our investment. And, you know, we went through thick and thin together, and if it hadn’t been for the support that we had from RRE and the iron stomach that you brought along with you in this partnership, I’m not sure we would have had the same outcome. So, thank you for everything.
Raju: That was fantastic.
Jason: Thank you so for joining, Matt.
Matt: Oh, my pleasure.
Will: Thank you for listening to RRE POV.
Raju: You can keep up with the latest on the podcast at @RRE on Twitter—or shall I say X—
Jason: —or rre.com, and on Apple Podcasts, Spotify, Google Podcasts—
Raju: —or wherever fine podcasts are distributed. We’ll see you next time.