Interviewee
Brian Gerkey is CO-Founder and CEO of Open Robotics, which drives the development and adoption of open software and hardware for robotics. Prior to Open Robotics, Brian worked at Willow Garage, SRI, Stanford, and USC. Brian is a strong believer in, frequent contributor to, and constant beneficiary of open source technology.
Transcript
Brian Gerkey: My name is Brian Gerkey. I'm with Open Robotics and what we do is we make open source software for the robotics industry. Primarily we're best known for two products. One is ROS, the robot operating system, and one is gazebo. The, which is a robot simulator. And ROS is used to develop autonomy applications for robots. So if you're going to build a robot and you need a lot of software to build the application for that robot, you want it to do something. Then ROS is the toolkit that you reach for to build that application. It has all the developer tools. It has the algorithms, it has all the device drivers and everything else that you need in order to build those applications. So if you think about like what runs on robots, it's ROS and then cause Zibo is a 3d physics based simulation that you can use to do all the testing and development in software only when you don't have access to the hardware or you want to do, for example, continuous integration testing of your algorithms.
You can do it all in a realistic environment. It just in software and these two platforms together, there are other things that we work on, but those two platforms together are what we're best known for. And they're used now across any kind of robot that you can imagine everywhere in the world.
Elisa Muñoz: Wow. Really interesting. Okay. So I, now how about an idea of what Open Robotics does, but like how did your interest in robotics begin? Like I know that you went to school for Computer Science, so how was the transition like?
Brian Gerkey: Yeah, so as when I was in undergrad at Tulane University, I was in a computer engineering program and there was a professor named Jen Jennings who had a lab and our department and it had robots in it. And so there was a time when I had, as an undergrad, I had an opportunity to, I was supposed to go find a lab and try to do something interesting in one of the labs with the grad students.
And he had robots and robots are really cool. So I just, you know, I just gravitate naturally toward that environment. And there's something about, and you hear this from people who work in robots pretty frequently, that there's something about writing code and then having it have a physical effect in the world that is really different from most software development, most software development. Ultimately you end up, you know, getting the right answer on a screen, but that's just not the same as getting something to move around in the world that you can observe with your own eyes. And I just found that really satisfying. So from there ended up going to grad school in a robotics lab and just been in robotics ever since.
Elisa Muñoz: What year are we talking about?
Brian Gerkey: Now you're asking me how old I am. This was a long time ago. This is the mid nineties when I first started working in robotics. And then I know I see by the lights, the whiteness of your eyes that I know it's, it robots have been around for a long time. So in, I started grad school in the late nineties, and that was, and that, that turned out to be a really important experience for a number of reasons. But if we think about how that flowed into open robotics, one of the problems that we had at the time was a problem that a lot of labs had. So if you were in a robotics lab, you would get these robots and your goal is to use those robots to do science.
You're supposed to do research with them. They're basically, you can think about them as like experimental instruments. It's like to an astronomer, you need a telescope to hear about us as you need a robot and you just need the robot to work, and then you can do experiments on top of it. But often the robots didn't come with very good software. The software that came with them was usually proprietary provided by the vendor and it didn't do everything you wanted and you couldn't change it. And so we are, you know, a bunch of confident computer scientists, so we'll, we can write code. So we just wrote our own software to replace what the vendors provided us.
And then, because we were kind of idealistic, we made it open source. And so we started a project that was called the player stage project in the late nineties. And we put it up on the internet for everybody to, you know, be able to work with. And that we started seeing, you know, from around the world, other labs started picking that up and using it on their own robots and extending it and modifying it. And so that's where I really got the interest in developing this infrastructure for robots. And that's really like, if you think about what we do as a company, we don't really build a lot of, we don't build robots ourselves. We don't even really build the autonomy applications themselves. What we do is I think of us as like we're filling a toolbox with tools so that everyone everywhere from a classroom to a research lab, to a product based company can take those tools and then use them to build the applications they want to build.
Elisa Muñoz: This is really interesting. Okay. And when you say “We”, how many people are you referring to? Like, how did you meet your co-founders?
Brian Gerkey: Sure. So I guess I have to explain a thing that happened in the middle. There was a company called Willow garage. This was a robotics R and D company that existed in Silicon valley from about 2008 to about 2013. And I ended up going to work there. And so I ended up running the software engineering team there, and we were doing a couple of different things. But one of the things that we were doing was really building this open source robot software platform that became Ross. And at the time the goal was to provide this platform to work with peers too. That was a particular research robot that we were building.
It was a large human size, two armed mobile robot that we were giving out to research labs. And we wanted to provide software to go with it. That's where Ross came from. So we were building Ross, it was work. That was all like a great experience. And we were starting to get a lot of traction. So we had labs and companies all around the world now starting to pick this up and contribute to it. So in 2012, there were a few of us at Willow garage who decided let's spin off from there and create our own company that has this particular focus. So rather than being a general R and D company like Willow garage was, we said, look, this, this open software platform, that's important enough on its own.
That's something that we feel passionate about. It's something that there's a large community out there that relies on all this software. And it really deserves to have a company that is dedicated to focusing on that. And so that's why that's when we created open robotics. And that was in fact, we just celebrated our 10 year anniversary this summer.
“I think of us as like we're filling a toolbox with tools so that everyone everywhere from a classroom to a ResearchLlab, to a product based company, can take those tools and then use them to build the applications they want to build.”
Elisa Muñoz: Well, congrats on that. I wanted to ask you what has been the biggest technical challenge while building ROS? What would you say was the hardest part?
Brian Gerkey: I'd say the hardest part historically, and it's still the case for us, is managing the breadth of applications that we're trying to address. So if you think about what does a robot, that's just like a, it's an old philosophical question, the robotics industry that everybody laughs at, because there's no clear answer, but it actually matters here because if we say we're writing software for robots, that means we're writing software that needs to work for a robot arm in a factory. It needs to work for a mobile robot. That's doing sidewalk delivery of packages. It needs to work for an autonomous car.
It needs to work for a free flying robot in the space station. There are NASA has robots in the space station that run ROS and has to work for drones. It has to work for underwater vehicles. And so the tech a really difficult technical challenges. How do you balance all that? How do you build a common platform that works well in all those situations, given the very different requirements that you have in all those situations. And so you have to, you have to its and it's still something we struggle with. We struggle with it today is how do we build something that's flexible enough to cover all those use cases and yet not so insanely complex to use and configure that nobody can figure it out.
Elisa Muñoz: Did you find any critical challenges, especially since I know that you're working mostly with software, but maybe in the hardware part, did you have any challenges when it comes to procurement?
Brian Gerkey: Yeah, because we focused on software. We actually, as an organization, I think we had it relatively easy during the pandemic. I would say that, and in fact it helped there, there were two things that really helped our experience through the pandemic. For one, we've been focused on open source and leading an open source community for our entire existence. And so we're, we're, we're very used to, you know, even outside of our own employees, our collaborators who work on projects with us are spread all over the world. And we collaborate with them asynchronously online all the time.
And we do all our work in public online all the time. So when everybody had to go home, it wasn't that different. The, the, the kind of work style. And then we, even within our team starting several years ago, had already been adding folks to our team who were full-time remote because, you know, we're headquartered in the bay area in California. We have an office in Singapore as well, but not everybody is willing to relocate to one of those two places. There, there are great people out there who have great skills and they already live somewhere else. And so we had been, you know, hiring people who were working full-time remotely anyway. And so then in fact, I remember the day that we all had to go home from our Bay area office.
You know, some of, some of our people were kind of freaking out like, oh my God, what do we do? But they're their own colleagues on the team could just say, it's going to be okay, guys. I've been doing it, you've been working with me for the last five years and I've been working from my house the whole time it's going to be okay.
Elisa Muñoz: Thank you so much, Brian, do you have any last advice for future entrepreneurs? I mean, especially since you have tons of experience in the field.
Brian Gerkey: I'll just say that it's to pick something that you are truly passionate about. That's what, that's what I've done. That's what we've done with open robotics is we, we do it because we love it and it, if you're going to do it for a long time, you, you have to love it. There can be other benefits to it, but it's gotta be something that you really believe in.
Elisa Muñoz: Okay. Straight to the point. Thank you so much for being here and sharing your experiences with us. This was really interesting, Congratulations on the 10 year anniversary!
Brian Gerkey: Thanks for having me.
