The Manufacturing Executive
The Manufacturing Executive

Episode · 1 year ago

Perspectives on Automation, AI, & the Future w/ Andy Lonsberry


If the Wright Brothers lived in the 21st century, they might be the Lonsberry boys. Andy Lonsberry earned his Ph.D. by researching machine learning and artificial intelligence for bipedal walking robots. His brother focused on biologically spiking recurrent neural networks.

Together, they dived deep into allowing nonlinear systems to learn how to improve themselves. Then these two Ohio boys and their autonomous robot traveled to the west coast in search of funding. The trio came home with $15 million.

In this episode of The Manufacturing Executive, Andy Lonsberry, Founder and CEO at Path Robotics, an AI robotics software company, talks about how the midwest houses the future of robotics.

Here's what Andy and I discussed:

  1. How Path Robotics got its start
  2. Uncovering automation talent in the midwest
  3. How machine learning and AI are starting to change the manufacturing game

To ensure that you never miss an episode of The Manufacturing Executive, subscribe on Apple Podcasts, or Spotify, or here.

Listening on a desktop & can’t see the links? Just search for The Manufacturing Executive in your favorite podcast player.

It's a couple Ohio boys going to the West Coast. Didn't quite know how it was going to be received, but having an autonomous robot with us gave us a lot of credit. Welcome to the manufacturing executive podcast, where we explore the strategies and experiences that are driving mid size manufacturers forward. Here you'll discover new insights from passionate manufacturing leaders who have compelling stories to share about their successes and struggles, and you'll learn from B tob sales and marketing experts about how to apply actionable business development strategies inside your business. Let's get into the show. Welcome to another episode of the Manufacturing Executive podcast. I'm Joe Sullivan, your host and a cofounder of the Industrial Marketing Agency guerrilla seventy six. When you think of technology hubs in the US, where does your head go? Silicon Valley, Boston, Maybe Austin, Texas, but why not the Midwest? Well, today I'm talking with the founder of an AI robotics software company out of Columbus, Ohio, who put midwestern talent up against anyone, and our conversation you'll hear his story and also his perspective on where the AI and manufacturing automation revolution are taking the industry. Let's get into it. Andy Lawnsbury is CEO and Co founder of path robotics, Columbus Ohio based AI robotics software company. Well, working on their phds at Case Western Reserve University, he and his brother discovered a market need for industrial welding robotics. From there they founded path robotics, where they work to redefine and reenergize American manufacturing, and he earned his degree in mechanical engineering at the Ohio State University. Andy, welcome to the show. Thanks, Joe. Great to be here. So, Andy, before we get into it, I was wondering if you could just start things off by telling us a little bit about your personal journey and kind of how path robotics came to be. Yeah, not a problem. So half robotics started while I was doing my PhD. At the time I was focusing on machine learning and artificial intelligence for bipeed walking robots. My brother was focusing on biologically spiking recurrent networks. So in general, both of us really focusing on machine learning and allowing high degree or freedom on LADER systems learn how to improve themselves while we're doing our PhDs. We kind of always know that we wanted to start a company, but we just didn't know what to do. I think a lot of what we'd seen in academia was great research being done and sometimes people taking that great research and throwing it at a market and it never quite sticking right and it always felt us like it was a little bit like the cart was before the horse. It was this great piece of tech and not really a market understanding. And instead we wanted to flip that. We really wanted to find a great market pain point and they create a product that could help really address that pain point. So we didn't know what that pain point was. So we set off by building a consulting company, consulting firm, and we started exploring northeast Ohio's manufacturing market and the goal here was to not only be able to help people on a day to day with the consulting business, but eventually find a market pain point that was so big, based on so many customers that we were talking to, that would make sense to turn our consulting business into a real company building a product. So it sound a great to us at the beginning, where like this is awesome we're going to be able to figure this out on a couple months. It just didn't quite happen that fast. It was a long journey for us, about twenty four months of exploration. We're talking all sorts of different companies. We were looking on the record pressing, we were looking at medical automation, we were looking at transmissions, we're kind of looking at just a...

...lot of different markets. Are really trying to understand again where a pain point, or if there was a pain point in the manufacturing industry, that could really utilize our backgrounds in our skill set to help create a product solution for that. So time is going by, we haven't found really a market pain point yet and so of course there's that one faithful day. Well, we're at customers facility where there to talk about finite elmine analysis in the president of the company walks and basically says, let's not talk about that, let's talk about welding, and he goes on to tell me the story about how they have this big pain point and it all revolves around finished welding. They're having a hard time hiring. Well, there's they're having a hard time retaining, while there's any time they bring in somebody new that's never welded before. It takes some three to six months of apprenticeship to become a certified well there at their facility and it's really becoming this bottleneck that's not allowing them to grow. They had two options that they were looking at. One was offshoring potentially, but they're worried about losing quality. They weren't hitting the normal numbers you would need to really make offshoring worth it in terms of like seeing hundreds of thousands of the exact same part. They're a high mixed, low volume company. They are cashlow business, so they want to make what is ordered that day and they want to produce it. They don't want to make Tenzero, a hundred thousand of the exact same thing and put it on a shelf and sell it out throughout the year. If threezero different parts and so again, offshoring did it really quite make sense. Their other option was looking at robotics, and it sounds like a good thing, like industrial robots. You've seen what other people have been able to do with them. So they actually purchased to sell. They found out that classic industrial robots, though, aren't really viable for a company like them that have not super tight tolerances because everything's handmade. Again, they're making different parts every single day. They would have need custom fixtures, custom robot programs, they need a whole systems engineering team to really make an industrial robot work, and so there was never able to actually work for them. What they really wanted was just more human welders and that wasn't possible. So they posed us the question, could you replicate, with the robotic arm, a system that could represent, as close as possible, some of our human welders so that we could just be able to put in one of any of the stews into a three jaw chuck? You we do. Will press one button and the system would be able to scan, recognize on its own what it should be welding and determine how to weld it and then perform the welding process. That's when we kind of take a step back. We looked at the market. We found that this is a pain point that was felt not only by them but by many. We decided to make that jump and jump into the the welding industry and that's really when the journey kicked off. Of Well, now we have a pain point, how do we go make a solution? Happy to talk about that next part of the journey as well. Yeah, I just man, what was that moment like when when, all of a sudden you realize like wow, there's there is something here, there's a big hole and we might be able to fill it. Yeah, no, it felt what? It felt great because we had been explored for literally like twenty four months for that whole and it just kind of felt like it was something that connected so well with what my brother and I were doing for a research perspective. We needed something that could adapt, that can learn, that can use vision, that can use other inputs to make decisions on the fly. Again, that was something that was like really court are our technical backgrounds, and then to the fact that it had to do as welding. Both my brother and I, we've been kind of ingrained and manufacturing for a very long time. We've been welding since the age is of Eight, nine and ten. We've been making custom motorcycles for a long time. It just was something that just really spoke to us as well and just felt really natural. So it's super exciting to finally feel like hey, this is the market. This is what we're going to go to do. Let's...

...put it in stone and start marching forward towards it. That's awesome. How long ago was this? See, so I want to say this was in two thousand and sixteen, two thousand and seventeen. Okay, cool. So fill in the gap then, between front then and now. Yeah, so worked with that company. They gave us the initial funding to go create this system. My brother, myself and my other two cofounders, which was my father and another guy named macline, basically went to a basement. We carved out a small space in a basement of a foundry. I think it was roughly like toned Scraffee, so twenty by ten. We put two robots into that little space with us and we started building outs what would be the first, you know, fully autonomous robotic system. That was also a long process. Lusara, a lot of hurdles that came at us as we're trying to build. So a lot of technical challenges that came to your kind of surface. As again, making a system learn how to do something as challenging using vision as the main input and feedback is also not a not an easy system, easy easy task, but after grinding for quite a long time in that basement and living through some unique experiences of what it's like to really live in a basement. After months, we came to the surface. We had those two prototypes. We deployed them at the customers facility they run to today. Currently they're making you a hundreds of mufflers a day with the systems. And then, once we deployed them, we had to make a choice. Do we want to continue to grow organically or do we want to accelerate? I was pretty adamant that I wasn't really gonna go back to that basement and I want to accelerate our growth. So we we decided to fight at the Silicon Valley. We were taking a shot at the venture capital market. At first we're super concerned. It's a couple Ohio boys going to the West Coast. Didn't quite know how it was going to be received, but having an autonomous robot with us gave us a lot of credit and we were able to close that first piece of funding it rather quickly. Came back to Ohio with your two and a half million dollars for a first seed run seed round. Talk to a couple venture capitalists here in Columbus gathered some excitement very quickly around what we were doing, and then we close our twelve and a half million dollars series a, you know, a couple a couple months after the series seed, so at the end of two thousand and eighteen, to put some timeline on this, we close about fifteen million dollars in funding. We're very small company, just just a handful of people moved out to Columbus and really started building what is path robotics and started building out the company that we want to create, which is the system of from the product perspective, it's fully autonomous welding. It's smart, intuitive system that can learn, that can make decisions on its own and that you really utilize his vision as it's it's core input feedback. You know. Fast forward to today, we just announced our series be funding. It was led by addition where yearizing that capital to continue to expand into our vision and continue to expand into making robots more intelligent and more adaptable to their environments. Well, that's awesome. CONGRATS on the success, Andy. They so much show. So Andy, I'm a midwestern guy, born and raised in Milwaukee and I've lived my entire adult life in St Louis, where my business partner John and I have built our agency over the last fifteen years. And I've heard you talk about how you take pride in your midwestern roots as well, but also that you've uncovered so much talent in the Midwest. For you specifically it's been in Ohio. So just kind of curious if you talk about that, let's give give the Midwest a shout out here and just kind of hear your side of that story. Yeah, the Midwest is has been exceptional. So I I born raised in the Midwest, grew up in Ohio. Have family had a machine shop in Youngstown, Ohio. Actually Worn Ohio for a long time while I was younger. So really have just kind of been in grain and manufacturing and grain in the Midwest for my entire life. I think a big part of what this company is and who...

...this company is is represented by my midwest roots and by the Midwest roots in general. That we have a path and what we've seen to be just phenomenally successful for us is hiring people in the Midwest. When I started this company and after we got funded and funded from Silicon Valley. There's always that concern that we would need to move to Silicon Valley to really get the best talent, where we need to move to the east coast to be able to get the best and it's just not the case. It's not what we've seen at all. We've seen that the talent here in the Midwest is exceptional on the engineering side, on machine learning side, on the AI side. There's really really great people here near us and we are trying to cultivate and really go after that town source as much as possible. I think Pittsburgh is really also helped set off the town in the Midwest. They're known as a city for robotics right now. They're one of the fastest growing robotics scenes kind of in the world, and we're trying to have Columbus also get in there. But it's not just Ohio, it's not just Pennsylvania. We're seeing it kind of all over the place in the Midwest that the talent is real, it's here, it's in numbers and we're trying to cultivate as much as we possibly can from it. That's cool to hear. Yeah, and you probably is Ohio State Guy would want to admit it, but I know there's a ton of a tongue of greats of going on in Michigan as well. On the robotics side. We've got, you know, a client there and we've I just see a lot of a lot of companies in the robotics world popping up and Michigan and you know, throughout, but yeah, throughout the Midwest. I'm in the marketing industry, obviously, but there's a lot of talent here and, you know, costs of livings is great and it's you know, they're these cities are like advancing. There some great universities, like here in St Louis Is Washington University and St Louis University. That breed a lot of really smart young talent and it's we got to keep these people here in the Midwest, though. I think it's part of the challenge. So definitely totally agree. I think Washington and St Louis is an exceptional school. I actually think that's where our most recent fundraiser are fundraise was led by addition and the head of addition, Lee, I think he actually went there for his MBA. So he's got easy got roots in the Midwest as well. And Yeah, yeah, you know supports what we're doing, also because we're doing it here in the Midwest and you know, we're very close to our customer base as well, which is why we wanted to stay here. That's awesome, wash you happens to be my Alma Mater, so it's what brought me down to St Louis and kept me around. So awesome. Cool. Well, let's all right. Let's shift gears back to AI and robotics here. So you know, I've had a handful of guests on recently, Andy, that have come from one corner or another of the manufacturing automation world, and I was just I'd like to hear you give your perspective on how machine learning and AI are starting to change the game and will continue to change the manufacturing landscape. Yeah, of course. So again, my research really focused on machine learning and artificial intelligence. Steadd my cofounders, my brother, my cofounder, his research also really focus on so I feel like we have like a deep connection to the low level functionality of how machine learning really works and we feel very strongly that it's going to have a huge impact on all sectors that affect our daily lives, and manufacturing obviously being a big part of that. So manufacturing, for a long time there hasn't been a lot of intelligence systems deployed and manufacturing there's been some great machinery. It's come from you see, and C technology and whatnot, and that's really paved the road for manufacturing as to today. But I think that next step level up is going to come from machines learning from themselves and machines being able to take data based on what they've just accomplished and be able to tweak themselves completely, and I think that's the big, big change that we're going to see here going into the future, is how to enable machines again to make smart decisions on the fly and why it's so important, I think in manufacturing, especially the manufacturing that I focus on, which is again high mix, lower volume, whatever parts is running at...

...a skew of ten to a hundred or maybe even a skew of one, it's really, really tough to not have an intelligent operator or intelligence system be doing that operation. Maybe you really really takes usually some sort of operator interaction or skilled labor interaction to be able to do something like that. Most of the time, when you think about manufacturing at scale, you think about somebody that's making a hundred thousand to a million of the exact same thing, if at least that's the picture that kind of comes into my mind when I used to think about manufacturing, and it's just not the case. There's so much manufacturing from fabrication, your metal fabrication, to small runoff and that really is a very difficult thing for anyone to use any sort of automation in that sort of sense. And I think personally, to be able to move into utilizing automation and low volume runoffs, you're going to have to have a system that can adjust on the fly based on what it's seeing and based on what's asked of it at hand, and to me, that's the big impact that machine learning is going to help change the game on. It's a good perspective, you know. Is there anything anty you'd like to comment on, in particular about the welding industry and and what's what you see happening in the years ahead? Yeah, so I think right now the welding industry is it's going through a change, like a lot of other industries. Currently we're seeing this massive demand for welding. I think we're seeing massive demand from manufacturing in general right now. Covid drastically shows some weaknesses in our supply chain kind of across the board, and I think more emphasis again is coming back on local manufacturing, and just not in the United States alone. This is happening. I think I'm kind of all over the world where there's this. This is additional momentum around local manufacturing and for us the United States, you know, manufacturing for quite some time has been changing and evolving. A lot of it is gotten shipped off shores and again, as things broken our supply chain, this re vitalize the need for more local manufacturing. And welding is such a key piece of manufacturing. I think it's fifty percent of all man made products in the United States have some sort of welding process being done to them, and so we see this massive, massive need for more welders more welding in general, and again we're not seeing that same demand being matched by the number of workers entering the welding occupation and for us we're seeing this is again the reason why we're building what we're building. We're trying to help revitalize and reenergize and gifts as much support as we can to local manufacturing by giving them the tools that they need to be able to continue to scale and to continue to bring manufacturing back to the United States and to continue the harding the manufacturing world here in the United States. But again, also everywhere is there's a local manufacturing is just going to be continue to be, at think, a big issue for most people out throughout the world. I sound like a broken record when I say it, but every every podcast episode I have one way or another, you know, the lack of skilled workers and and labor shortage, it's affecting everybody, really appears to be the biggest challenge that American manufacturers are facing right now. Yeah, I would totally agree, at least from the companies that I talk to. I mean I talked to a company today. They had it just pass on a ten million dollar order because they didn't have the capacity to be able to support that order, and then they're really looking to us to help them make that order happen. And so we see it again. Everyone that we're talking to is running into this issue and they want to be able to grow, they want to be able to sustain the momentum they're getting and their feeling from their customer base and was well thank being just such a...

...vital part of that manufacturing process and we see this kind of happening day and day out right now. Bad believe it. Andy, I heard you recently use the acronym rass, or robot as a service. I was wondering if you could talk about what that is and how you see this concept maybe starting to take root in the manufacturing sector right now. Yeah, so robots as a service, I think, is a newer term that's been around for maybe a couple years, but certainly not decades by any by any means, and the whole concept of robots is and service is that classic and industrial equipment is usually purchase on, you know, a capital expenditure and it's usually a onetime up front cost to pay for that piece of equipment that would get paid back over years. That is been a long standing practice for them, especially in the manufacturing environment, and I think large to your one people again work cash isn't a problem to go get and buy equipment. They might be standing by this practice for a long time in the future. But what we're trying to do is enable small to medium size manufacturers to be able to utilize robots immediately, and utilizing them as a service allows us to deploy robots quickly effectively and we get to take the burden on that. The system either works are you don't pay for it. And so how the system works, that how at robots as a service works for us, is there's no cap expend you sign up to a subscription for our robotic system. We deliver the system, we tell you the quality metric, we tell you the kpis around cycle time that you're going to see on your parts and that system has to perform every single day or you don't pay. And so what we're trying to do is enable these manufacturers to be able to move into automation immediately. There is no large CAPEX spend up front. They don't have to go try to get the cash to make that happen. They are looking at some thing that's very comparable to their labor prices and their labor costs that they can put onto their opex spend and immediately be utilizing robots to continue to expand their manufacture operations on day one and not three years down the line, and they again they're going to see that return or that return on their spend immediately. It's not a three or two or two or three your payback period. It's an instant in the first hour of our systems employed, our customers making money based on on our subscription. It feels like a fantastic model to me. I mean it seems like you guys are emulating what what happens a lot in the software industry and it only makes sense that, especially we're, like you said, we're companies, are used to dealing with these large capital expenditures and now you can put these things on site immediately without the huge upfront cost and some of the risk is is on you rather than them, which is, you know, kind of lowers a barrier to entry. So smart way to go. I'm curious to see how it plays out in the years ahead. And thanks, Joe. And that's a big thing that that you touched on. Their the risk. We're trying to take on as much risk as we possibly can. We've seen and we've talked to a ton of customers that have purchased robots and they were sold a pipe dream when they bought that robots, but they still paid for it and then she'll try and make it work. We try to reduce that risk to zero. If the system doesn't behave like we tell you it's going to behave, from being able to weld your parts, being able to hit the quality of metrics that you need, be able its cycle time that you want and desire, you just don't pay and we come and take it back and you have no risk of losing cash to see if something like this would work. It feels like a no brainer to me. Thanks trip. Well Andy, this was a great conversation. I really appreciate you doing this today and would love for you to tell our audience about where can they get in touch with you and learn more about path robotics. As Jeff so to get... touch with us, please go to wwat paths roboticscom. That's our new website. You can see the equipment that we're we're currently selling. You can sign up for a demo to come see the system live for yourself and understand how this can work for you in the future. Beautiful. Well Andy, thanks again for doing this today and, as for the rest of you, I hope to catch you on the next episode of the Manufacturing Executive. You've been listening to the manufacturing executive podcast. To ensure that you never miss an episode, subscribe to the show in your favorite podcast player. If you'd like to learn more about industrial marketing and sales strategy, you'll find an ever expanding collection of articles, videos, guides and tools specifically for bedb manufacturers at Gorilla Seventy sixcom learn thank you so much for listening. Until next time,.

In-Stream Audio Search


Search across all episodes within this podcast

Episodes (130)