Long-Term Investing Ideas in a Volatile Market
Simon recently spoke with a $35 billion global asset manager about how they're navigating the market volatility. The key takeaways are to think long term, tune out the noise...
7investing lead advisor Steve Symington sits down with Lior Elazary, CEO of inVia Robotics, to discuss the accelerating demand and commercialization for industrial robotics.
July 6, 2020 – By Simon Erickson
Robotics innovation has already improved many industries for the better, effectively streamlining everything from surgeries to manufacturing, e-commerce fulfillment, and home care.
But we believe the robotics industry is only just getting started as tech companies around the world strive to unlock the full potential of this lucrative niche.
In an exclusive interview with 7investing, inVia Robotics co-founder and CEO Lior Elazary discusses the competitive landscape of the robotics industry, the challenges of differentiation and commercializing robotics at scale, and the benefits of implementing the budding concept of robotics-as-a-service.
Lior holds a Master’s Degree in Computer Science with an emphasis in artificial intelligence. He co-founded and successfully exited three tech companies prior to launching inVia Robotics, a subscription-based robotics company that provides smart-warehouse solutions to e-commerce customers.
0:00 – Introduction, background on Lior, the founding of inVia, and early regulatory/commercialization challenges for various robotics business models.
2:12 – Solving warehouse problems using robotics for e-commerce customers.
6:09 – How inVia helps mid-sized businesses compete with Amazon Robotics/Fulfillment by Amazon, differentiation & cost-reduction by adapting to existing warehouses.
9:50 – Offering customers an end-to-end robotics solution, using AI software to help humans and robots work in harmony.
12:02 – Embracing robotics-as-a-service as the future of the robotics industry.
16:20 – Challenges of building new robots, the importance of machine-learning software to expanding use cases for existing robotics solutions.
18:58 – The challenge of building new companies, pursuing organic growth versus growth by acquisition.
20:50 – The disadvantages of having deep pockets; how too much cash can encourage R&D but discourage commercialization.
Publicly-traded companies mentioned in this interview include Alphabet, Amazon, and Softbank. 7investing’s advisors and/or guests may have positions in the companies that are mentioned.
This interview was originally recorded on June 18, 2020 and was first published on July 6, 2020.
Hi everyone 7investing Lead Advisor Steve Symington here. And I can’t tell you how excited I am to introduce my guest today. He’s Lior Elazary, co-founder and CEO of inVia robotics. Lior, welcome. Thanks for coming on.
No problem. Thank you very much, Steve, for having me on the show.
So part of the reason I’m so happy to have you here is that we share similar backgrounds. I previously worked as a software engineer, I was implementing neural networks and machine learning algorithms for object detection and feature extraction from satellite and aerial imagery. And investing in artificial intelligence for the last several years is one of my primary areas of expertise. I understand you have a master’s degree in computer science with a focus on artificial intelligence. And you met your co founders of inVia while attending a Ph. D. program in robotics at USC.
Yeah, that’s right. I was actually working on the PhD for a while. About seven years, almost finished it and I actually dropped out to start this company and they give me the Masters as a consolation prize.
Beautiful. So tell me a little bit about the genesis of your company, then I understand you initially mulled the idea of building home robots for elderly care but pivoted when you recognized the need of the e-commerce industry for better fulfillment solutions?
Yeah, so you know, our first idea was to build a robot that will operate at the home. And we’ve looked at various methods and various things that, you know, from what technology will allow us to do. At USC, we worked a lot on DARPA-related projects. You know, that kind of allowed us to really understand what robots can and cannot do. And one of the difficult parts we actually ran into was the business model. That we were going to be able to probably get a robot to maybe do 20% of the tasks at the home, but we might have to charge around $50K to $60K for it. And for most people that’s just way too expensive, especially for elderly care and stuff like that. I still would love to solve that problem. But I think that problem – it’s just that there’s…it’s not just technological. It’s also there’s a lot of other regulations and rules and things like that, you know, have to be done.
It’s really challenging. Yeah. Exactly. On any meaningful scale to commercialize something like that.
Definitely, yeah. And we knew that, you know, maybe a few people will buy the robot because it’s cool. We were able to go maybe through doors, and our biggest task was cooking, which means, you know, pulling out a microwavable food out of the freezer and putting it in the microwave. So that was a really, we were really proud of it. We actually have a very, very early on video of that. But it’s, you know, again, just the cost of it was just going to be way too much. And right now, you know, you could get people to do a lot more on 120% of the work for a lot cheaper. So as we were building these robots, we were ordering all these parts from Amazon. So first of all, you know, I think this is one thing that people don’t realize in being able to get items as quickly as possible. It’s not just, we can order the little stuff from China as quickly as possible. It also allowed us to iterate and build robots very, very quickly. Because we’re able to get a part the next day and be able to test it, see if it didn’t work and get the next part the next day in the past would take months to do that. And so development time has increased significantly. One of the things we were ordering from Amazon were these aluminum sheets, like eight foot by four foot aluminum sheets. They would come individually wrapped. And we were like, how does Amazon do this? I don’t even make money off this and be able to, you know, send it Prime overnight. So we started looking at this stuff. And I previously had another company called EdgeCast that we co founded. And I talked to some of the customers over there. So the EdgeCast was a CDN — content distribution network. And what we were doing there is basically I think, what e-commerce is doing today, which is you have servers that are all over the world to be able to produce content as quickly as possible. In fact, we always joke around that we always have to be beat the speed of light because the speed of light is just too slow for people to get content. And now I joke around that my job is to get atoms faster than the speed of light. So forget about data, because you know, they’re not going to get the shipment fast enough. So I was going to a bunch of e-com customers, and all of them had the same sort of problem, which is, you know, one is people don’t want to pay for this work, right? You basically when you click “Add to cart,” you have a personal shopper that’s doing the work for you. And nobody wants to pay for that. And people want it for free. On top of it, they want it really fast. They want it now. So that’s putting a lot of pressure on our customers, our potential customers right now that are doing all the e-com work. And what’s happening is that their labor that they’re trying to get for that it’s not only doesn’t really exist, is it people just don’t want to do this work. So they actually don’t last long, but most of the customers we’ve talked to had about 100% turnover in their staff. People can only walk so long for so long. It’s a lot of miles. Yeah, just it was crazy. So they told us like, please, please help us. What can you guys do? And this is where we came up with a concept of the inVia system. I’m sure you’ve seen our robots, but what they’re able to do is they’re able to go in and solve the picking problem. So there’s no more, in our world, there’s no more pickers. There’s nobody running around in the warehouse, the people are stationary and they actually do what they do best, which is fine manipulation. So we go inside a tote or packet, they will do the stuff which is extremely complex right now. And they don’t have to walk so we completely eliminate the walking for them. And that makes a huge difference. Now, we’re also able to add additional tasks, such as replenishment, cycle counting, and that’s the sort of the beauty of the system is that we have one robot that is capable of doing several tasks, and really helping our customers. Big value-add there. Exactly, yeah. And that’s one of the things we’ve concentrated a lot on is, you know, how do we go beyond just picking how do we solve the whole supply chain end to end, especially just inside the warehouse. And that’s what our system had been able to do. And we’re very proud of it.
Great. So you mentioned Amazon when I first became aware of your company that was the first thing that came to mind was Amazon’s acquisition of Kiva Systems. That was early 2012. Yeah. And they, I believe, renamed it Amazon Robotics in 2015. And now, you can correct me if I’m wrong, but I think they just they essentially use their Kiva Systems technology, that warehouse technology to bolster the value proposition of their Fulfillment by Amazon service. So to what extent do you compete with Fulfillment by Amazon and Amazon Robotics?
So we’re actually sort of direct competition with them, right, and we will offer our robots to Amazon-type of facilities. Uh huh. Now what happens though, with the Kiva system, and Kiva really kind of shook things up a bit because, you know, Amazon sort of acquired them and kept them to themselves, but there was one difficult challenge with the Kiva system, which was to adapt into existing warehouses. Yeah. That’s something we concentrated a lot on. So with the Kiva system, often it’s really becomes a construction project. And it takes a while to do it. And if you look at Amazon right now, I mean, they got about, I think, almost two, oh gosh, over 200,000 robots running around, but only about 30% of the warehouses are automated. And our reason for it is that they’re not able to bring the Kiva systems into existing warehouses. So that’s one of the challenges that we have to solve is how do we get our system to operate in an existing infrastructure without having to change the warehouse? And by solving that we actually opened the door to everybody who’s trying to compete against Amazon and might not be at the scale of buying a robotics company. Yeah. We become the robotics partner. And as a result, they don’t have to spend massive amount of capital because often all that capital is spent on you know, re-engineering the whole warehouse, redoing the floors and everything. Now they just get our robots deployed into their existing system, they’re able to up and run basically right away. So we’re able to help all the like the mid-size businesses, who are sort of struggling, right, because they don’t have the capital or the expertise to go with the automation. But now they can actually afford automation to be competitive with Amazon at the same type of rates that, you know, they’re able to achieve with their automation but without that initial cost, and that’s what you know, we’re actually really excited about it. Because when we looked at the market, only 20% of the warehouses were automated. And because of that, because there’s not that many companies who are going to spend $150 million dollars to redo their warehouse. Now, the 80% remain un-automated and you just, you know, you wouldn’t imagine like what it actually looks like. And now by us being able to address that market, we really opened up the market to a lot bigger set of customers and really helping them not just compete against Amazon but actually survive in this world.
Yeah. And I could see a lot of those customers really being hesitant to say, “Okay, I’m just gonna let Amazon handle this.” It’s sort of like feeding the beast they’re competing with, and really, that that sort of answers one of the questions I was going to ask you next is, you know, as far as differentiators go, is that you think what you’re most proud of? Are there other things about your technology that really sets you apart? I’d say that’s that’s a huge one is the ability to adapt to existing warehouses because we see that a lot in the enterprise software world. It’s reducing switching costs, or adopting new technology is the biggest preventer. So what what else would you say you’re proud as far as differentiators?
Yeah, so I think there’s a couple of things that when we develop this, so we, like I mentioned a little bit before we didn’t just develop a picking solution, we actually developed a full end-to-end solution. One of the reasons we did that is that we noticed as we’re going to these warehouses, they were really missing a lot of the intelligent component of routing things inside the warehouse. So our system is actually composed of an intelligent software we call inVia logic that not only manages the robots, but also manages to labor to efficiently tell people, where should they be at different points in time, which is the same algorithms we actually use for the robots. But together as a harmony, they’re actually able to operate together. And it’s really interesting because people are not really good at repetitive tasks and doing that over and over again and staying consistent. But they’re really good at problem solving, and they’re really good at bursting. And what we’re doing with that system is being able to handle this sort of workforce that is able to burst and problem solve with a workforce that doesn’t burst very consistent, stays over and over which is the robots, but actually is not able to problem-solve. Whenever the robots run into a problem. They need a person to go and solve that problem for them. I think this is something that people don’t actually realize that’s happening. These are tools that help us do a really good job and a repetitive job over and over again. But we’re nowhere near the intelligence that even the most basic person has. Yeah. And what we do often is, you know, we call now the people, instead of pickers, instead of them running around in the warehouse, we call on robot wranglers, because now what they’re doing is they’re managing the robots. Yeah. But in turn, they’re able to pick at a very, very large rate, right. So yeah, instead of doing a few orders an hour, they’re doing now thousands of orders an hour, but without them actually physically doing it. They’re managing the robots. So the job has sort of transformed. And most of the people that have been doing it are actually really happy with that. It’s a much more fulfilling job for a person to do. You know, to sort of guide the robots right and tell them what to do. As opposed to, you know, them running around like a robot.
Yeah, I remember reading statistics on the number of miles those poor pickers were walking a day. I mean, really physically demanding, you think, “Oh, it’s exercise,” but a certain point, it’s just burdensome. Now that that kind of brings me to the challenge of commercializing robotic solutions on a meaningful scale. Now, I read some on your website, about the prospect, the solutions that you offer for robotics as a service solutions. Now that helps address the challenge of commercialization for you. And it seems like it’s a more attractive prospect for your customer so they’re not, you know, shelling out $50,000 a robot. But both from a creator’s perspective and the standpoint of that robotics consumer, tell me a little bit more about your thoughts on robotics-as-a-service, how you embrace it?
Sure. So we actually are very, very proud at calling our company, you know, truly doing it as robotics-as-a-service. Which we feel like it’s not so much rental, like some companies will just rent you a robot, you know, per month. We actually don’t rent a robot, we rent a service, which is the productivity of the robots. And we think that is really, really important because when you as a consumer, rent a robot, you’re in charge of its productivity. If it doesn’t perform to its, you know, expectations or what you want, you still have to pay for it every single month. And that actually is what the industry was used to before. And we’ve seen that happen many times where a bunch of companies will, you know, rent a bunch of robots, try to use them. It didn’t really work, they’ll push them to the side. And they’re not robotic experts. They don’t really know how to use them. In our world, doing it truly as robotics-as-a-service. We are not leasing the robot, but we are basically giving them productivity that they’re paying for. So if our robots are not as productive, it is on us to make sure that they are productive. Now we are as a robotics company, we know what we’re doing. We know how to make these robots productive. So and our customers, really what they’re interested in is the productivity. They want to know how many orders can I get out the door, right away? They don’t really care about the nuances of how many robots do they need? We take care of that. So that becomes a really good synergy between the companies where we’re aligning the incentives, and it really grows everything. And to your point earlier, getting that out to market is a really key point because that allows us to expand very quickly and be able to supply robots to as many customers as possible. Where in the past, because the customer had to take all the risk, that was very daunting to them, right? If you’re again, you’re spending $150 million on a project, you’re taking all the risk. If that doesn’t work — and we’ve talked to a customer who actually did that, and they said, you know, it took five years to complete. And it didn’t materialize to what they thought it would. And it was actually five years behind our technology. So it was just, you know, it’s a big spend and they took all that risk, we are actually taking that risk for the customer to make sure that the robots are going to be productive. And that alleviates them. And in fact, from an ROI perspective, it becomes almost infinite, because they’re saving from day one, the minute the robots are hitting the ground, they’re saving money right away, and they’re able to also grow with the system, as opposed to having you know, one of the reasons that designed that system was to project 10 years into the future, and figure out what system they will need, because it’s very difficult to expand. Now, today with today’s world, you know, like, we basically turn 180 degrees overnight, right with everything that’s happening, and companies are no longer that only does not have the ability to predict it, and it was very difficult to predict 10 years anyways, but now it’s even moreso just predicting til next year, what’s going to happen? Yeah, and then really has opened up the door actually to more deployments on our end, because more customers are willing to move forward now, because they know that, you know, they basically are using the robots for what they need today. And when they need something in the future, they know they can expand or adapt, or we actually have a customer said, “You know what, I’m going to move warehouses.” They know they can pack it out and move to a different warehouse, and do that.
Yeah, that’s fantastic. I mean, really, and you’re taking the words out of my mouth, over and over again, which is kind of nice. From an ROI perspective, it’s so easy for a customer to be able to say, “Okay, this is my potential return on investment.” And on the flip side, that makes it easier for you as a robotics company to say, you know, here is how much money we can expect from this and help everything pay for itself as you strive to move toward, in some cases, companies moving toward sustained profitability or investing toward that growth. Now, speaking of investing back into growth: Do you see, as you continue to gain traction, can you see inVia eventually building out its repertoire of robots beyond its core fulfillment niche, to different robotics solutions?
So we actually see our software being able to do that. So since our software, the inVia logic is able to handle a lot of tasks — and we actually do that right now, for example, interface with conveyor belt, and auto baggers and things like that — what we’re able to now do as well is take on some other robots that are in the system. So for example, there’s a bunch of pallet moving robots. And if we need that function in the warehouse, our system is able to guide that robot to again work in harmony across the whole warehouse. So we see definitely doing that. Initially, when we started the company, we didn’t want to build our own robots. We actually wanted to just do software and do the intelligence part. However, I think we ran into a huge cost factor, where you know, some of the robots were costing $20K-$30K. And we said, there’s no way we could, you know, do that because you need a bunch of robots to do a lot of work. So that’s when we ended up designing our own robot. And in fact, this is where AI really comes into play, which I think people don’t realize it. What we’re able to do is take a low-cost set of hardware, so off-the-shelf pieces of hardware, basically bent aluminum sheets that are not very precise, so they don’t cost a lot, and compensate them using AI. So instead of building a machine that is very costly, that is very precise, mechanically, we’re building a machine that is not as precise mechanically, by using vision systems, we close the loop to basically have a very precise machine so we can still go in and very precisely within millimeters pick up a tote. But we’re not doing it because our machine is very precise. It’s because it’s closing the loop like a person. We have a vision system, that kind compensates for that.
I see the system behind you and the camera and the sensors and everything. It’s funny because it’s deceivingly simple, but I think the real power is in the artificial intelligence that’s contained in that software system. So this, I guess this is potentially really interesting to me. Now you’ve co-founded and served as the chief architect of several businesses. You mentioned EdgeCast earlier. And I believe that was sold to Verizon in 2013, you said? And then you had HostPro that’s now web.com. And HostPro was acquired by Micron in the late 1990s.
Yep, 1999, right before the crash.
Oh, goodness, KnowledgeBase as well, was another one that I think was acquired by Talisma in 2005?
Yeah, that was a CRM.
So, solid track record so far, but dare I ask whether you could see inVia growing into a compelling acquisition candidate itself in the coming years, or is that something that’s on your radar? Or do you want to grow this organically and see what it becomes?
So I’d love to grow this. I mean, we do have investors that would probably would take on — even now what’s happening, we’re actually experiencing a tremendous amount of growth. Now, you know, probably bringing in more. Now, I would like to grow it as much as possible. I’m sort of in a comfortable situation where I do want to grow it now. Robotics has always been my passion, always wanted to do something robotic systems. And like I mentioned before, we always run into the business model. Always a problem. And I think this is why we don’t see the robots running all the way around, like, you know, we promised in I think, 1960s and 70s, right, that by now we would have robots running around everywhere. And it wasn’t so much the technology, but as much as the business models, just, you know, was very, very difficult to justify. And, you know, I would love to live in a world right, where we’re basically out doing the things that we want to do, and having robots doing all the stuff that we don’t want to do.
Those menial tasks that we don’t want to handle would would be fantastic. I actually saw I think a press release yesterday from Boston Dynamics, you know, they make the the rover, the running and flipping robots that they always have the viral videos. I think they said they’re selling their little dog, spot robot or whatever that is to any company that wants to buy it for $74,000. Ouch.
I see that. Yeah, that’s sort of the point right? Boston Dynamics is a really interesting case. They’ve been really helping with research right? Doing a lot of — companies have been pumping — I think now they’re owned by SoftBank — so they’ve been pumping a lot of money to do that research. But finding an application, right, an application that the robot dog can do that justifies $74,000 is very difficult. And that’s something that I think Boston Dynamics in the past had actually a problem with commercialization. I think they always tried to do this stuff, I mean their robots are great, right? They’re jumping, doing flips and things like that. But trying to find a commercialization is always difficult.
Yeah, and that’s part of the reason — for anyone who’s paying attention to these various transactions and mergers and divestments — that I believe Google acquired Boston Dynamics at one point along with half a dozen other robotics companies, and subsequently divested them. And I think their goal was to create sort of delivery robots that could bring packages to the door from their autonomous vehicles. And I mean, it’s just one of those things where, you know, even Google with all of its cash is having trouble commercializing something like that. So that really makes what you are doing all the more impressive. So it’s just a really exciting technology and company, I’ll be watching.
Yeah, I think to your point actually, sometimes cash could actually work against you because you end up doing a lot more research. So we have to commercialize. I mean, you know, because we don’t have the bank account that Google has. So we got to draw income, and as a result, we might need to make sure that our customers are paying us money and as a result, we’re making sure that our technology fits what the customers want. And that I think that sort of sometimes plays into that
Sort of created a necessity situation — the “mother of all invention” correct?
That’s really all I have for you. I’m really happy that you came on, and I’ll be watching your company. I really appreciate your time. Thank you so much for joining me today Lior.
No problem. Thank you very much for having me. I really had a lot of good time and if you’ve got any questions or anything like that would be happy to answer.
Of course. And for those of you listening, you can learn more about inVia at their website. It’s inViarobotics.com. That’s i-n-v-i-a robotics.com. And thank you for tuning in. We are here to empower you to invest in your future. We are 7investing.
Simon recently spoke with a $35 billion global asset manager about how they're navigating the market volatility. The key takeaways are to think long term, tune out the noise...
Anirban and Matthew were joined by Alex Morris, creator of the TSOH Investment Research Service, to look at seven former market darlings that have taken severe dives from...
On episode 5 of No Limit, Krzysztof won’t let politics stand in the way of a good discussion - among many other topics!