EP_026 - Kevin Dowling of Kaarta

Speaker A: Hello good people. You're listening to Startup Berg, the podcast dedicated to highlighting the Steel City's growing startup community. I'm your host, Adam Simone. Let's get started. All right, well, welcome to episode 23. This is take two. Uh, we've got kind uh, of a fun episode today with Kevin Dowling, the CEO of Carta. Um, Kevin, welcome to the show.

Speaker B: Thank you. It's good to be here.

Speaker A: Yeah, thanks. I think what you're working on is really neat. I'd love to get a little background on yourself and Carta, uh, what you guys do.

Speaker B: Sure. Um, I first came to Pittsburgh back in the uh, uh, early 80s, um, as an undergraduate in applied mathematics. At that time CMU didn't have an undergraduate computer science program. And that was the unheard of. That. Yes, that's right. Computer, uh, science itself was a field looking for a home. But it eventually created its own home in the school of Computer Science at co Carnegie Mellon. Um, I became a staff member after graduating uh, and working closely with uh, Takeo Konadi and Hans Moravec and later Red Whitaker and we formed the field Robotics center there at Carnegie Mellon.

Speaker A: That's incredible.

Speaker B: Grew up in that lab. Um, I finished my master's and PhD at um, Carnegie Mellon in robotics. And in 97 when I finished my degree, left uh, for back to New England where I grew up.

Speaker A: Oh, where in New England?

Speaker B: Uh, Connecticut, um, which is where I grew up but ended up working outside of Boston.

Speaker A: Oh yeah, I lived in uh, the Norwich Mystic M area, Southeast Connecticut and went to school in Boston.

Speaker B: Okay. Yeah, yeah, so I enjoyed New England a great deal. Uh, we spent about 15 years there. Um, I went through a succession of companies, startups, uh, mid sized companies, learned a great deal. Uh, probably the most high profile success there was a company called Color Kinetics which was actually founded by two CMU uh, people, George Mueller and IHOR Liss. Uh, we created a wonderful company that went public, got acquired by Philips for around 800 million in 2007 and then uh, I stayed on during a lockup period and then joined another startup doing conformal electronics. And we created ultra thin electronics that could go on the body or inside the body. Uh, I left there after four years of leading the R and D team and commercializing that technology. Um, I then came to Pittsburgh actually at the invitation of Henry uh, Thorne, who co founded or founded Athon and four Moms here in Pittsburgh. And so I then headed up the uh, uh, engineering group at four Moms, uh, for a few years.

Speaker A: How was it to come back to Pittsburgh.

Speaker B: It was great, actually. It was incredibly impressive how much Pittsburgh had changed for the better. And I'd continued to involve in the alumni side at Carnegie Mellon. So I was here several times a year for those obligations. And you could see everything from bike trails to restaurants, the cultural scene, and importantly the startup scene. It had changed dramatically from the late 90s.

Speaker A: And what do you attribute that change to?

Speaker B: I think a mindset to actually create and live here, uh, in Pittsburgh, but on the part of many entrepreneurs that they didn't necessarily have to go to the west coast to find everything. It's still challenging, I think, in terms of access to capital and attracting and retaining talent, but all of that's improving very rapidly. And, and I will tell you, having spent the last year, I travel to the west coast frequently. I've been, it's really almost a second home. Uh, every five to six weeks I'm there and I talk to investors. And the profile of Pittsburgh in that region is very high right now. I think Uber and the new Argo investment by Ford.

Speaker A: That's right.

Speaker B: Incredible. And many, many of the people in the valley, in the robotics startups are in fact CMU graduates. And so if you look at Chris Urmson at Google and Bill Lee at, uh, Knightscope Robotics and many others, there's so many examp of Carnegie Mellon people in the tech scene.

Speaker A: It almost feels like this tipping point for Pittsburgh where it's really starting to accelerate and gain this incredible momentum.

Speaker B: Yep, I think that's exactly right. And one of the things I had left four moms in October of 20, uh, 14 or 2015. Sorry.

Speaker A: Okay.

Speaker B: And, uh, I formed the Pittsburgh Robotics Network, which is trying to connect all of the different robotics companies here in Pittsburgh.

Speaker A: Um, and why did you do that?

Speaker B: Because I found that many of these companies didn't know what other companies were doing or other companies weren't there. And then students at Carnegie Mellon. Carnegie Mellon is graduating about 120 graduate level students, masters and PhD students every year just in robotics. And they weren't being well served by the community in the sense that they didn't know about what was going on here. So, uh, since that time, since we formed the Pittsburgh Robotics Network, we've gone on to have meet and greets with students who now say, boy, I had no idea there was so much going on here in Pittsburgh. And they now have options. They don't have to go to California and spend five to six thousand a month here.

Speaker A: And they get the benefits of this great city and the Economics of that, um, and we retain that talent. That's awesome. So, uh, Pittsburgh Robotics Network, is that you, um, still involved with that organization?

Speaker B: Yes, very, very much. Okay, great. Uh, but we've fortunately been able to bring aboard other people who are other local entrepreneurs, uh, because it sounds like

Speaker A: you're pretty busy right now.

Speaker B: I am, I am. But fortunately, there's a group that's helping to lead and guide the Pittsburgh Robotics Network. Now, Jackie Erickson, who's, uh, helped us communications, she's now with the New ARM Institute.

Speaker A: Uh, and how do people find, uh, the robotics network?

Speaker B: On Twitter. We're robopgh. R O B O P G H. Great.

Speaker A: Okay, cool. Well, I hope anyone listening who's into robotics and doesn't know about that goes find you guys and, uh, and hooks into the network. Um, okay, so, um, you, uh, Left4Moms, you started the Pittsburgh Robotics Network 2015, and you're starting to think about, um, what's next for you. Right?

Speaker B: Exactly. And, uh, I met with Sanjeev, uh, Singh, who is a local entrepreneur. He's done a number of small, uh, companies here in Pittsburgh, and, uh, has been a professor at Carnegie Mellon for many years. And, uh, we talked about something that he had going on in his company, which is called Near Earth Autonomy. Uh, Near Earth Autonomy is about autonomous flying missions.

Speaker A: That's right.

Speaker B: And they fly everything from drones to full size helicopters using their control and sensing systems. Um, Sanjeev had a piece, uh, of really interesting research done by one of his graduate students, Ji Zhang. And Xi had evolved that significantly from the Carnegie Mellon work to work, uh, at Near Earth Autonomy. And Sanjeev had already started to carve out that piece and create, uh, a company around it. We saw that the technology had a lot of value, not just in robotics and autonomous vehicles, but in other areas. And, uh, I was very excited about it because it was the best performing type of system I'd ever seen that helped to manage, uh, your localization and mapping of environments.

Speaker A: Okay, so, um, can you explain what that means? Localization and mapping environments. And what does it mean to develop a technology that performs well in that arena?

Speaker B: Sure. So there's, uh, an acronym called SLAM in Robotics that was posed 30 years ago, and it's about simultaneous localization and mapping. And this basically answers two questions. Where am I? And what's around me? But it has to answer both of those simultaneously. It's a very difficult sort of chicken and egg problem.

Speaker A: Yeah, I ask myself that every time I wake up in the morning.

Speaker B: Similar, but your sensing is quite good.

Speaker A: Thank you.

Speaker B: Uh, but to teach a robot where it is and what's around it is very challenging. And through a combination of different sensors, cameras, lidar laser, radar, and an imu, we can now create.

Speaker A: What's an imu?

Speaker B: Uh, inertial Measurement Unit. It's basically a device that, through motion, can determine its path, but it actually drifts quite a bit, so you need other information in order to anchor it. Uh, but the combination of these types of sensors, plus some core algorithms, allows you now to build a very faithful, uh, map of the surroundings and determine your position within it, all without having any prior maps, any infrastructure, or using gps. Uh, you don't need any of that.

Speaker A: So this is the core technology that, uh, Ge Zhang developed at cmu. Correct. Um, that anchors what you're doing at Carta.

Speaker B: Correct. And then he completely reinvented it for the commercialization side, really rewrote it. Actually improved the accuracy significantly.

Speaker A: Okay.

Speaker B: And as proof of that, uh, last year about this time, we won the Microsoft Indoor Localization Competition in Vienna.

Speaker A: Congratulations.

Speaker B: Thank you. Uh, it was, um, a nice honor. And then secondly, if you look online in the kitty benchmarks, uh, these are run by the Karlsruhe Institute of Technology in Germany. These are online benchmarks where you take data sets, process them, and then tell them with the path and the localization within that. And we're number one and number two on that list over the past year and a half.

Speaker A: It's not a bad place to hold.

Speaker B: Definitely.

Speaker A: Okay. So, I mean, you saw this technology and you're like, this is something special. What's the commercial application for this? How did you see an opportunity to really spin a company off?

Speaker B: So even to preface that, one of the hard things for a lot of technology entrepreneurs to understand is that you can't fall in love with your technology, no matter how cool it is, because it skews your vision of what the commercial possibilities are. Just because you have a better widget doesn't mean you're going to have a market for that widget and people willing to pay for that widget. Uh, so my main focus initially was to look at what is the business that we create around this. And we discovered that we had three good options, one of which had already just been started, which is a SAS portal. Basically, you can upload your own raw sensor data and we can convert that into a map for you. And it's on our website@carta.com. carta, by the way, is spelled K double A R T A.

Speaker A: Right.

Speaker B: Uh, so K A A R T a dot com we have a SAS portal in which people for free can upload their raw data sets and we'll process them. If they start to use very large data sets or many data sets, we will charge them. And there's a very uh, economical fee that they are um, assessed on, which has to do with how much data they're processing.

Speaker A: Are people actively using that portal?

Speaker B: Yes, daily. Uh, really. And then uh, secondly we saw an opportunity to license the technology again, a nice high margin opportunity to take the core technology, provide it in executable form so that people put, could put it on robots or other devices, uh, even their own hardware systems to do mapping and localization. Then thirdly, we've created and built and sold uh, systems that uh, are hardware based, that incorporate LiDAR, that use cameras and have an IMU and a processor board. But these are uh, primarily off the shelf devices. So there's very little that's customized to use our algorithms. So we're using a commercial off the shelf processing board. Uh, the processing power is not augmented by an FPGA or an ASIC or any or gpu. This is all purely uh, off the shelf hardware. And so these handheld devices you can now walk around with or mount them to a vehicle and it will map and localize for you.

Speaker A: So who uses that kind of technology?

Speaker B: Uh, robotics people, autonomous vehicle people. And so we have customers in both categories already using this and licensing our technology.

Speaker A: That's great. Okay, so you had, um, so you said the first thing you wanted to do was to look at that commercial viability and figure out what kind of business you could build. And there seems to be a lot of different, um, areas for you to approach. So um, uh, tell me about your team. How many people do you have, who's on it and how did you decide what to attack first?

Speaker B: Sure. So the team, uh, right now. Ji Zhang has Recently defended his PhD.

Speaker A: Congratulations.

Speaker B: Joined Carta full time as of March 1st just a few days ago. And uh, we also have uh, uh, Kathy Pattison, who I worked with at Color, uh, Kinetics, the LED company that I was part of for many years. Uh, she's in the Boston area. And then Wade Sheen, who's a very experienced both CEO in the past, uh, raising money, but is focusing on sales and business development right now. He's very, very familiar with doing business, uh, around the world, um, currently today as we speak, he's in Asia.

Speaker A: Um, you must have customers around the globe.

Speaker B: We do. In fact, we have more customers overseas than we do right here in The US right now, um, because the sort of early adopters who want this are focused on large urban areas where the density of people and the density of buildings is now um, a very attractive prospect for mapping, which is the other main aspect of what we do. So it's not just robotics, it's also uh, surveying, mapping and building as built models of, of buildings.

Speaker A: That's a big part of your business.

Speaker B: That's correct.

Speaker A: And plan moving forward.

Speaker B: Yep. And the way I would distinguish those two markets is that autonomous vehicles right now is a very promising market. It hasn't emerged as a revenue generating market. There are massive investments going into that, but very little in the way of revenue being generated from autonomous vehicles.

Speaker A: Makes sense. However, early in that curve.

Speaker B: That's correct. But construction worldwide is a $4 trillion market of that there's uh, on the order of 50 to $100 billion in surveying, mapping, assessment as built surveys and so forth. And that's where we're playing right now. And that's where our revenue is really being generated through.

Speaker A: Okay, so you have this near term, um, kind of build your business foundation and then invest in other technologies that can enable longer term further out markets to develop.

Speaker B: That's right. We're fortunate that this technology can be used in a wide variety of areas. But that's a tricky situation for a startup, uh, because you need to focus on something that will grow your business. Absolutely be an attractive niche. But surveying and mapping is. If you look at companies like Trimble or Garmin Magellan, other folks who are in the geospatial business, and even those in the lidar business, um, this is a very attractive market.

Speaker A: I was on your website this morning and there's a video that shows a map of some outdoor space. Uh, looks like it might be somewhere in Europe, uh, potentially. Um, very cool, like interactive. And that's the kind of maps that your technology can develop.

Speaker B: So think of something like the original Star Trek's tricorder. It was a recording device, a processing device and an output device. That's why it was called that at the time. Uh, these devices uh, could provide assessments of people in space. So it was very science fiction y but very good. This essentially is providing that capability as you walk through a space in real time. You're seeing a map emerge in front of you, um, emerging m directly on the screen and you can interact with that, you can rotate it, get better viewpoints, you get answers at the point of work. And for interior spaces, we believe that the inside is as big as all the outside. The idea of just knowing your position in the world or scanning outside is limited. In fact, we can do much greater things on the inside of buildings where today people use measuring tapes and clipboards. There's transcription errors, there's uh, uh, a variety of errors that happen through calculations and other things. Uh, you can use high end LIDAR and build maps, but it's not real time. It requires a lot of post processing. Our systems require no post processing at all.

Speaker A: So what really, is there any. What sets you apart from a technology side? Is there something defensible, patentable? Um, how do you build a wall around this?

Speaker B: Yeah. So we are, uh, heavily patenting, uh, these ideas and while there is certainly patents that are happening at the technology level, uh, we're also focusing on applications and we're also focusing on uh, uh, products. So we're sort of a tiered approach to patenting. So we have a very strategic approach to doing that.

Speaker A: Makes sense.

Speaker B: I did that at Color, uh, Kinetics, a previous employer, and that ended up being worth between 5 and $600 million just from the patenting activities.

Speaker A: Wow. Okay. Yeah, there's a lot of value there, right?

Speaker B: That's right.

Speaker A: Um, okay, so what do you see for uh, for Carta in the next 12 months? What do you hope to, where do you hope to be in the year?

Speaker B: So we have about 8 to 10 distributors worldwide right now. We want to expand that. Uh, so uh, Wade and the team are working hard to uh, enable that. That will act as our, essentially our amplifier to get out to more customers. Because each of those distributors knows their region, they know that business and they know their customers in that region much better than we could do by building a direct sales force.

Speaker A: Uh, makes a lot of sense.

Speaker B: That's a big part of it. Marketing too. We just rebranded the company. We're positioning in a different way. We're now going to trade shows which we had done, uh, very little last year. Um, but we did well last year in terms of revenue, in terms of margin, uh, dollars coming out of that. This year we're going to scale that much further.

Speaker A: Okay, are you growing the team as well as you scale those?

Speaker B: Yes, that's right.

Speaker A: And you're growing those in Pittsburgh, Boston,

Speaker B: uh, mostly in Pittsburgh. Uh, but we are a worldwide company. So uh, we have people now on the west coast, we have people on the east coast. So we are a worldwide company, but our, uh, core operations are right here.

Speaker A: And what kind of profiles are you going to look to add?

Speaker B: Oh, technology, of course. Um, we're looking to Continue to augment the software team. Uh, primarily on the hardware side. We're working with, uh, outside folks to actually do product development. Um, I don't think it makes sense for us right now to build a long term large engineering team that has mechanical, electrical, and optical and so forth. Uh, it does make sense for us to focus on the software side since that's where the real value is. That's right. And our product, uh, portfolio right now consists of, uh, three hardware products. And we're, uh, going to expand that in the future. But, uh, we don't necessarily have to, uh, build a team. So here in Pittsburgh, there are companies that are already doing product development, uh, for a fee. So we work with folks like that. Um, uh, so that's the technology side, the sales side, the marketing side. Um, I'm looking to also bring someone aboard in the finance side at, uh, some point. Now we're working with outsourcing for some of these activities already.

Speaker A: Well, hopefully you make so much money you need to accelerate that.

Speaker B: Yes, yes. It's a good problem to have.

Speaker A: Yeah, it is. So if people want to follow your progress or talk with you about the Pittsburgh Robotics Network or what you're doing at Carta, is there a way to connect with you?

Speaker B: Sure, just Kevanarta.com.

Speaker A: wonderful. And that's K A A K A R T A.

Speaker B: Correct.

Speaker A: All right, great. Uh, okay. Uh, you know, I have a couple important questions for you. I always end on, uh, number one, is there another entrepreneur in town that, um, you think our listeners should hear from? I mean, you've been around this community for a number of years, you've gone to other communities, and you're back in Pittsburgh. So obviously there's something really special happening here. Um, and I just want to put as big a spotlight on all those things as I can.

Speaker B: Um, there's so many good ones. Uh, I think a very interesting one might be to go back to Hans, uh, Moravec at Sigrid. He founded the company or co founded the company. He's still with the company. Uh, he's a pioneer in artificial intelligence, mobile robots vision. He's done a really remarkable set of things. Uh, Sanjeev Singh for his work in autonomy and flying systems.

Speaker A: Absolutely.

Speaker B: He's done many companies. Um, you've spoken to Jurgen and Jang and others. Uh, but, uh, there are many.

Speaker A: Yeah, there are. There are a lot. Well, I love the idea of Hans, Moravec and Sanjay, Sam, and I'll have to get them on at some point.

Speaker B: That would be great.

Speaker A: And, uh, question number two favorite restaurant in Pittsburgh.

Speaker B: So I don't often go to sort of high end restaurants, you know, the Mount Washington. They're good food and so forth. But I really like sort of classic Pittsburgh places. And one of my favorites, um, m. Mainly for nostalgic reasons, but I still go there occasionally is big gyms in the run. That's great because we've had many great times there. Yeah, I was there recently with dinner with Red Whitaker and a few other folks in the area.

Speaker A: That's awesome. All right, well, Big Jim's great. So we've got, uh, um, Kevin Dowling, Carta. Ah, thanks for the time today. Tell us about what you're doing.

Speaker B: Oh, you're welcome. Thanks for the opportunity to tell you about.