Cesium CEO Patrick Cozzi on the Geospatial Gurus Podcast

Highlights:

• Career path
• Cesium as a brand name
• Open standards
• Partnerships with Komatsu and Intel
• Cesium for Unreal
• Open-source and 10 years of CesiumJS
• Trends in 3D geospatial
• Autonomous driving
• Leadership and culture at Cesium

Full Transcript:

Girish Joshi: Hello, and welcome back to another episode of Geospatial Gurus, where we strive to bring top leaders on the hot seat and chat with them about the science of geography and their take on how it helps human society and world economy. As you know, our guests are spread over many industries, technologies, geographies and subject matter expertise, but our format is always the same, share some knowledge with a whole lot of fun.

A renowned engineer by profession, an author and contributor to 10 books on mapping, computer graphics and gaming, our guest today brings a whole new dimension to geospatial data. Not satisfied with bringing 3D visualization of satellites in space, he created the Cesium WebGL engine and 3D Tiles, an Open Geospatial Consortium Community standard for screaming massive heterogeneous 3D geospatial datasets. He is co-creator of gLTF, the open standard 3D runtime asset format that has been adopted by Microsoft, Google, Oculus, and Nvidia amongst others.

And if you love Christmas, then you love Santa Claus and NORAD Tracks Santa’s journey using CesiumJS across the globe each year. Patrick Cozzi is the CEO of Cesium, a Philadelphia based software company that is building an open platform for 3D geospatial data. Patrick, welcome to Geospatial Gurus.

Patrick Cozzi (PJC): Hello. Thank you for having me and thank you for that warm introduction.

Girish Joshi (GJ): My pleasure, Patrick. Before we begin, I want to ask you something that has been nagging me at the back of my head. May I?

PJC: Sure.

GJ: A boxer, a biker and a techie. Tell me you like Jim’s on South Street too and we can plan a day and reschedule this podcast.

PJC: Yeah. Being Philadelphia based, I certainly do love my cheesesteaks and I would say I’m definitely a techie. I started programming when I was in eighth grade and I did a lot of game development, self-taught throughout high school. And I guess I was an interesting kid because in addition to programming, I would also train at the boxing gym for years, maybe even a decade or two. So I’d be hitting the speed bag and I would also be thinking about my Turbo Pascal code. But I learned a lot. I mean, besides staying in shape, I learned a lot through boxing in terms of quick feedback, discipline, and continuous learning. And then on the motorcycle side of the house, yeah, actually a lot of the folks at the boxing gym had motorcycles, right? And that got me interested. It wasn’t until a little later in life that I got a motorcycle and now I try to ride about a thousand miles a year, and I did about a thousand this past season. The western suburbs of Philadelphia have some beautiful roads. It’s great riding. I have a mid-sized cruiser.

GJ: So I ride a Ducati Monster S2R as well as I have an off-roader over here in India, the Royal Enfield Himalayan. We might just have to meet up some day and see if we can ride a few hundred miles together.

PJC: I would love that. I don’t know that I can keep up with the Ducati. Those are extremely fast, but I would love to go out for a ride for a day.

PJC: Sure. So for me personally, I think like many kids of my day, I was really interested in video games and ultimately, how do you make video games. And that’s what got me into programming in high school and then computer science in college. But when I was an undergrad, I actually spent most of my internships and coursework more on systems programming, like operating system type work. But when I graduated, I decided that I wanted to come back to Philadelphia and I was looking at companies coming to the career fair and they’re alphabetically listed. And it happened that the very first company was Analytical Graphics Incorporated. I said, well, wow, look at that. I could come back and live where I’d like to live and get back into computer graphics. So I joined AGI, who’s doing analysis and visualization software in the aerospace and defense industry.

PJC: Yeah. It’s a great question. And it actually goes back to our heritage there, starting at AGI with aerospace visualization, right? So when we started building Cesium, we thought that we were building a visualization engine for aerospace. It was only after we made it open-source that we saw all these other markets that really needed a 3D geospatial platform. For our heritage in aerospace, we really had to focus—and still to this day—on time-dynamic, high precision, global scale 3D visualization. That time dynamic-piece, if you think about satellites moving in time and space, is really important, and Cesium is one of the elements in the atomic clock. So that is where the name comes from.

PJC: Sure. Open standards are very dear to my heart and I think kind of central to our ability to move forward. When we look at producers and consumers, we see so much 3D geospatial data that’s available today on the producer side. When we look at consumers, we see so many different vertical markets that once you get insights and have experiences from this 3D geospatial data, well, what we need in the middle there is that platform to bridge that gap and to make it easy. And we see so much opportunity and so much potential, but also so much work ahead of us. No one vendor is going to do it all and we need collaboration, we need innovation, and we need to form a community to create that ecosystem. Open standards are at the center of that collaboration and that interoperability.

PJC: Sure. So we’re just so excited to see just the amount of inbound interest around 3D geospatial coming from corners of the world that we didn’t even know had an interest in 3D geospatial. I mean, I think back to the earliest days of Cesium where we still thought we were building a visualization engine for aerospace, and then suddenly along comes Red Bull, who wants to use Cesium to visualize the X Alps adventure race. And wow, that really opens up your eyes, right? So we love to work with partners and vertical solutions to help them realize the potential of 3D geospatial. Specifically with Komatsu, they’re in heavy machinery for earthworks, so think bulldozers, dump trucks, excavators, and they are very progressive and forward-looking and their vision for smart construction, where they’re bringing technology to the construction site to make it more efficient, more safe, and more accurate.

And they do anything from having a CAD model available to their machine so that when the excavator goes to put the shovel down, it can’t go below the target surface, but they also do a lot of data capture; drones, rovers, the machines themselves have GPS on them, and Cesium is the software provider for the visualization of that digital twin of the construction site gathered from all of those heterogeneous sources and gathered over time so that we can understand the progress of that construction site. We’re very excited to be working there. And when we look at their vision, our vision is very similar for the general 3D geospatial platform. In addition to Komatsu, correct, we are working with Intel. It’s really exciting for me. So I was actually an intern at Intel after my freshman year in undergrad.

So about 20 years later, it’s really amazing for me to be able to collaborate with Intel now, in this new capacity helping them serve the enterprise market for geospatial.

PJC: So a lot of my personal thesis and the thesis for Cesium as a company is that the problems that we need to solve to advance 3D geospatial, many of which have been solved in the computer graphics industry, whether it be games, movies, scientific visualization, and we apply our knowledge of computer graphics to 3D geospatial. This relationship here between Cesium and Epic, where we’re bringing Cesium visualization– global scale, high precision, highly accurate, WGS84 ellipsoid using 3D Tiles, into Unreal Engine? It has amazing visuals. In addition with the latest version of Unreal Engine, where they’re creating a micropolygon renderer and dynamic global illumination engine, there’s opportunity to combine everything that a game engine brings with everything that Cesium 3D Tiles brings for real-world 3D content. It’s just an opportunity that we couldn’t pass up.

And what we’ve seen is just unbelievable inbound. Since we’ve announced this last year, we’ve seen so many markets, especially in the modeling and simulation industry just having an unbelievable amount of interest. And I think we’re about to hit kind of a new era in 3D geospatial as we bring this to market.

PJC: Yeah. So you’re right. It’s been about 10 years since the first line of code for Cesium, and Cesium’s history is in many ways aligned with the WebGL history. When WebGL was a ratified 1.0 standard making 3D available in web browsers, Cesium development started right around that time. So it’s been an amazing ten-year journey, right from open source project, to product, to business unit, to now company, to now foundation for ecosystem. As for your question on balancing the open source community with creating a business, that is always a key question around open-source. How do you make it sustainable? How do you make it scalable? How do you create and nurture community while you’re also building a business to help fund the contributors? So for Cesium, we use what’s called an open core business model where we provide useful standalone software in the form of CesiumJS that you can use on its own, it’s open-source ,free, and that’s for web-based 3D visualization.

And then we provide optional, but complimentary value add software subscriptions. This is in the form of Cesium ion, the SaaS platform that provides 3D geospatial content such as global terrain and 3D buildings, as well as pipelines for bringing in your own raw data and then optimizing it as 3D Tiles. So in this way, we can provide an open community that is truly using open-source and then have the opportunity for those that are interested and potentially would benefit to provide our value add software. Then in addition, we do partnerships in verticals as we discussed before to really help folks who are leading or are very innovative in a particular market. We combine our technology with their market access and their market expertise. And it’s usually just an amazing relationship that helps us just create value quicker.

PJC: That’s a great question. And I think as you alluded to there, when you look at the value chain from data acquisition and data processing to data dissemination to end user insight, over time, the value shifts, right? Parts of that value change commoditize and then parts of it become where you can add unique value. And we are seeing just a flood of data acquisition with anything from new space to open data policies to crowdsourcing. And that’s really what has driven a demand for Cesium to now make it really easy to have these building blocks to then serve particular markets. As for markets where we see a lot of interest and utilizing 3D geospatial data to create solutions, certainly in modelling and simulation, I think there’s a bit of a Renaissance that’s happening there, in addition, construction broadly, earthworks, AEC included for the built world, and then also aerospace. Cesium is shared across 19 different markets and aerospace continues to be in the top two.

PJC: Yeah, I’m a big fan of the vision of electric and autonomous vehicles. And I think that over time, we will get to a high level of autonomy. And meanwhile, the autonomy providing kind of driver assistance, I think is a really fantastic incremental way to get there. What I’m excited about is when you look at say a particular autonomous vehicle, well, it has LiDAR, sonar, radar. It has all these different types of sensors that are being used for perception to then help guide the navigation of the vehicle. And that data has so many potential uses, right? Even outside of autonomous driving in terms of building a digital twin of the world, a time dynamic digital twin of the world. And Cesium’s works thus far in the autonomous driving world has always been around visualizing the data that is collected, whether it be LiDAR point cloud or even very fine-grained data from the machine. And I think there’s much more we can do there bringing it to the digital twin vision.

PJC: It’s a great question. I feel that I was always lucky that when I was learning programming in high school, I found a lot of books and I read the books and I really kind of wanted to give back. And then likewise, when I was in undergrad and in grad school, I felt that I had some great mentors and internships and I wanted to figure out ways to give back. So when you look at my teaching and my writing and how that’s influenced Cesium, one, just on our messaging, we’re not salespeople. There’s not a single quota carrying sales person at Cesium and we believe in being an authentic authority and sharing the knowledge and the opportunity and the potential of 3D geospatial, and that’s how we do business. Within Cesium, there’s a huge focus on mentorship and many of our teammates, they come right out of undergrad or grad school and Cesium is a great place to learn and that’s how we want to build the company.

PJC: Yeah. I have several tattoos. I’m not sure which one you’re referring to, I have two tattoos on my arm of my father when he was in the Korean War. I had some very cool photos, they’re probably from about 1951, and about 10 years ago, I had a friend do the artwork and tattoo. I’d love to get more tattoos. I’ve been so busy right now that I might have to wait a few more years. Ultimately, I’d actually like to do my full back. I’d like to do a back piece that would be the history of computer graphics.

GJ: That was the one I was referring to. I have my sources, Patrick.

PJC: I’m impressed!

GJ: At some point, you will have to share your tattoo artist’s name with me because I have been waiting for the longest time to get two anagrams done with both my daughters names on each arm.

PJC: Very cool.

PJC: I mean, I think a question or maybe a set of questions just for the industry as a whole is if we pick a time period, maybe five years out, what do we think that we’re going to achieve that today we don’t think is possible? And what do we think we’re not going to achieve that seems like we should be able to today? I think we always surprise ourselves.

PJC: I would ask that the whole community, right? You get insights from all different perspectives and folks working in different areas.

PJC: Well, I’m a big fan of your conferences and I’m very much looking forward to both conferences. Thank you for the invite.

GJ: Thank you so much for your kind words, Patrick, and more importantly for your time today, sharing some of your thoughts, how you started in this industry, why are you passionate about geospatial data, your leadership style, as well as what you envision as trends and opportunities for the future. I wish you nothing but the best and I hope to meet you real soon.

PJC: Thank you. Thank you for having me today.

GJ: Thanks, Patrick. Have a wonderful day.