Keenan Wyrobek, "Drones + Robots + Africa"


Roboticist Keenan Wyrobek’s first career was founding Willow Garage which created ROS (Robot Operating System) that now makes robotic components inter-operable. After creating a foundation to sustain this vital piece of open source software, Keenan went in search of new engineering challenges. He became intrigued with the idea of helping developing countries deliver crucial medical supplies to areas that are poorly served by roads. He co-founded Zipline, the VC funded venture that is improving the lives of people in rural Rwanda. Zipline is now poised for profitable growth. Listen to this inspiring and brilliant engineer speak passionately about solving major problems while building a business.

Click here to read full episode transcript. 

Topics covered:

  • Sal’s Pitch for Angel Invest Boston Syndicates
  • Sal’s Introduction of Keenan Wyrobek
  • How Keenan Wyrobek Became an Engineer
  • Keenan Wyrobek Starts Willow Garage & Creates ROS (Robot Operating Systems) the Ubiquitous Open Source Software that Makes Robot Components Inter-Operable
  • Robotics: 90% Software & 10% Hardware?
  • Keenan Wyrobek Stumbles upon the Idea of Zipline
  • “And I think the reason we got that traction in our conversations is that this was a really visceral problem. Everybody's one degree away from it. It's either your grandmother or you or your sister has gone to the hospital and the doctor says, "I could have helped you, but I don't have X."”
  • It Took Two and a Half Years to Refine the Solution for Delivering Perishable Medical Supplies with Drones
  • “Our first plane had a wingspan that was probably about four feet and what we operate now is 10 feet. So it's went from a plane you could carry under your arm with no problem to a thing that's a beast.”
  • VCs Saw a Huge Opportunity
  • In the Midst of the Drone Craze in 2014 & 2015, Zipline Was Quietly Making Progress
  • New Version of the Drone Is Built to Scale
  • Keenan Wyrobek’s Advice to Young People Interested in Drones
  • Parting Words

Transcript of "Drones + Robots + Africa" at TEDMED

Guest: Roboticist & Founder Keenan Wyrobek

Sal’s Pitch for Angel Invest Boston Syndicates

SAL DAHER: Hi, this is Sal Daher of the Angel Invest Boston podcast. If you've been listening, you might have noticed that I love being an angel investor in Boston. The reason for this is that there's so much going on in the startup space here in Boston. Practical founders working with leading inventors, venture capitalists, angel investors, patent attorneys, it's a really exciting scene. Now you can join us in syndicates, which allow people who are not part of the angel investment community to invest alongside Boston's leading angels. I invite you to leave your email address at in the syndicate section and we'll be back in touch with you to help walk you through the qualification process as an accredited investor. Remember, there is no obligation to invest when you put your email address there. I hope you really enjoy today's podcast.

Sal’s Introduction of Keenan Wyrobek

SAL DAHER: Welcome to Angel Invest Boston. Conversations with founders and angels, usually in Boston. But today we're coming from beautiful, sunny, Palm Springs, California where we are at the TEDMED 2017 conference. And this is a very special episode. We are going to be interviewing Keenan Wyrobek and I can say a few words to you, drones, robotics, delivery of perishable medical supplies in Africa, saving lives, making money. A guy who seven years building open source robotics software and then goes to Africa to solve a problem of expiring medical supplies in warehouses and people out in the country suffering and dying for lack of medications. So this is an unbelievably exciting episode and I'm very grateful that Keenan Wyrobek is here sitting down with us.

KEENAN WYROBEK: It's great to be here.


How Keenan Wyrobek Became an Engineer

SAL DAHER: Awesome, Keenan. So tell us the story. The reason I asked this is that it's really important. Somewhere out there there's a 12 year old or a 15 year old Keenan Wyrobek who might emulate your model and do something amazing the way that you've done. So tell us how you got to be to the point where you started thinking about doing a startup. Don't tell me about the startup, tell me what led you to that point.

KEENAN WYROBEK: I've always been passionate about building things that could help people. I started in medical devices and I really enjoyed it. I really enjoyed the interaction of nuts and bolts engineering with the understanding of biology.

SAL DAHER: So, you were trained as an engineer?

KEENAN WYROBEK: Yeah. I was trained as an engineer, always loved engineering. And my first exposure to really what I would call hard core inventing was as an undergrad at Johns Hopkins. I worked in a lab that was the entire basement floor of the urology department. And this lab had literally 20 machine tools.

SAL DAHER: Which department?

KEENAN WYROBEK: The urology department. This was a lab with 20 full sized machine tools run by this brilliant designer, staffed by five professional Romanian trained machinists who were just master craftsmen. And basically we spent all of our time working for the head of surgery who was upstairs in the urology department. And anytime he had some idea for how he could do a surgery that was impossible using some tool or some gadget, we'd make it for him. And he'd go upstairs and test it. That was just awesome to me to see that process of somebody who really understood what was needed in the OR in this case. And then people who were just brilliant at taking that need and making something and trying something.

KEENAN WYROBEK: And of course, making those kind of medical devices, that level of hand craftsmanship and things and this level of prototypes, that were often being used on humans, upstairs, was just awesome to see that happen.

SAL DAHER: Excellent.

KEENAN WYROBEK: So I thought that it was going to be my path. I came to California for grad school, to Stanford and I started to get cynical quickly.

SAL DAHER: Grad school in engineering?


SAL DAHER: Mechanical engineering?

KEENAN WYROBEK: By then I was really focused on robotics.

SAL DAHER: Robotics, okay.

KEENAN WYROBEK: Originally inspired by medical robotics, the likes of the Da Vinci robot. And I actually had the opportunity to use one of those in a pig lab at Johns Hopkins and see what it was like to have your head inside of a body in 3D and have everything magnified 10 times so that vessels that were a millimeter looked like a garden hose. It was so cool. That inspired me into robotics and really bringing together the disciplines of electrical and mechanical software, engineering and systems engineering.

KEENAN WYROBEK: At Stanford, I took a hard right after getting disillusioned by the world of med devices and medical robotics in terms of ... It just felt like the incentives were ...

SAL DAHER: So you never worked in a medical device company? Just in the academic setting.

KEENAN WYROBEK: Only as a consultant. Doing consulting projects for a few different companies. And tried to pursue getting some concepts prototype patented. I'll never forget this one conversation I had with this med device I was very proud of. It was a surgical instrument and the guy who was this outside investor and advisor to Hopkins looked at this thing and he said, "I could probably sell that for $50." And that's not a business for me, right? It's too simple.


Keenan Wyrobek Starts Willow Garage & Creates ROS (Robot Operating Systems) the Ubiquitous Open Source Software that Makes Robot Components Inter-Operable

KEENAN WYROBEK: That started me getting disillusioned with how medical technology targeted at the US market kind of gets born. I took a hard right and went into robotics but way outside of medicine. I started a project called Willow Garage. We built something call ROS, Robot Operating System, which has become the Linux of robotics. Many companies have founded it. We're actually celebrating its 10th anniversary next week and it's been used by many companies in the Boston area, use it. Rethink Robotics bases their stuff on ROS now. And lots of other startups now build their platforms on ROS.

SAL DAHER: Oh wow.

KEENAN WYROBEK: And so I spent seven years doing that and had a great time. Building open source community was fantastic, took a long time to find the funder for it but once we found it, now there's a foundation that's a steward of that open source community. And after that, I wanted to find something that I could get really excited about. And I had learned from other startups that I was involved with, I was one of the founders at Neato Robotics and a few other places along the way, I wanted to find something that really could scale. Neato Robotics, it did okay, right? They make the little vacuuming robot and it's doing fine but it did not have the transformative effect that I was hoping for.

SAL DAHER: By the way, I have a Rumba in my beach house and I tell you, I would have paid $500 even if it didn't clean at all. Because my grandson gets such a kick out of the thing, running around from the robot and seeing it go on and so forth.

KEENAN WYROBEK: That's great.

SAL DAHER: But anyway, please, before I digress.

KEENAN WYROBEK: Yeah, so after the open source robotics foundation was formed and leaders from Willow went on to be stewards of the ROS platform, I was trying to figure out what I wanted to do next. I did a bunch of things along the way to figure that out. I built a learn trade app for kids that's quite cool and innovative, but nothing I thought was going to be a business. I worked on a lot of different robotics companies, companies that wanted to be robotics companies and I would take the time to help them understand that there's no such thing as a robotics company. Find your customer, understand their need, there's a chance it's a robot that meets that need, but there's also a 90% chance it's not a robot. And I did that for a while.

SAL DAHER: While you're passionately hoping that it is a robot so you can build a robot.

KEENAN WYROBEK: Yeah, I mean I like the challenge of robotic technology.

Robotics: 90% Software & 10% Hardware?

SAL DAHER: Is it true that in robotics it's 90% software, only 10% hardware? I've heard this said.

KEENAN WYROBEK: It just depends what you're doing. It just depends what you're doing. I mean what's exciting today is you can buy super capable robot arms, for example, off the shelf and then yeah, you can do a company that does nothing but software. That wasn't true 10 years ago. So there's opportunities for entirely software centric robotics companies.

SAL DAHER: So it's becoming truer and truer?

KEENAN WYROBEK: Absolutely. You could look at it the other way too, which is one of our goals with ROS and one of the reasons it's so powerful today is it's enabled specialization. Think of it like a middleware that plugs everything together. So the reason you can go buy an arm off the shelf and just turn it on and write some code and go is because it's got ROS integration. And the same with the sensors, they have ROS integration.

SAL DAHER: Gosh there's all this depth that I didn't get during the talk.

KEENAN WYROBEK: This is my past life.

SAL DAHER: I know but it's fascinating. This is tremendous. So please continue.


Keenan Wyrobek Stumbles upon the Idea of Zipline

KEENAN WYROBEK: Sure. Sure. So coming out of Willow I had credibility as a robot guy and so people reached out to me asking for help. And like I was saying, that's where I spent a lot of time helping people not make robot companies. And got to know my co-founder at Zipline. And we really bonded over this notion of how to think about a product and how to think about customers. And we got really excited about not making anything until we actually knew who the customer was, in some abstract sense, but we could shake their hand and they told us they wanted to buy it, that kind of thing. So that we didn't end up going down the rabbit hole of convincing ourself that our idea was going to work and go build it first and see if it worked later.

KEENAN WYROBEK: And we actually explored a number of different concepts. Some in the manufacturing space, all over the place. We had heard from some people in the developing world, healthcare world, who told us about, "Hey, look, this logistics thing is a big deal." So we got out into the world and really studied it and got to know people in Central America and in Africa who were running these health systems and these doctors working in these places.


KEENAN WYROBEK: And it just clicked. We saw, well, warehouses of expired medicine, which was a big ingredient for us because we were really worried, okay maybe we can solve a logistics problem, but was upstream supply the problem now? So once we saw the warehouses of the medicine that was expired I was like, "Okay, there's supply."

SAL DAHER: Okay. I spent 30 years in emerging markets and this is a classic thing. They have the crop but they can't get it to market. That's the problem.


“And I think the reason we got that traction in our conversations is that this was a really visceral problem. Everybody's one degree away from it. It's either your grandmother or you or your sister has gone to the hospital and the doctor says, "I could have helped you, but I don't have X."”

KEENAN WYROBEK: Exactly. And so within four months we had meetings with presidents of countries off of a deck that included some sketches of this idea that we could do this with drones and solve this big problem for you. And I think the reason we got that traction in our conversations is that this was a really visceral problem. Everybody's one degree away from it. It's either your grandmother or you or your sister has gone to the hospital and the doctor says, "I could have helped you, but I don't have X." That was the signal we were looking for, governments that were asking us how quickly we could do this and saying that they would pay for it, which was very exciting for us because we wanted something that we could scale. And then we got to work.


It Took Two and a Half Years to Refine the Solution for Delivering Perishable Medical Supplies with Drones

SAL DAHER: Get into the nitty gritty of the narrative here with what happened? Did you go out and devise exactly the right solution the first time out?

KEENAN WYROBEK: Not even close. Not even close.

SAL DAHER: We're fans of pivots in this podcast.

KEENAN WYROBEK: Sure. It was a pretty wild ride. As we got to know these countries better, we got more and more data from them on how their health system worked, their geographies, where the paved roads ended relative to the hospitals and clinics to help us understand how far to fly. And we were sure, and I still laugh at myself of how sure we were. Because we were sure, if we could serve a service radius of 20 kilometers, that that would be great. 20 kilometers would be perfect.

SAL DAHER: The thought was that you'd have these drones delivering critical supplies to hospitals where it's hard to get roads and so forth and it would solve all these problems with expiring medications sitting in a warehouse?

KEENAN WYROBEK: Exactly. Exactly. So the basic idea, drone on a paved road ... I'm sorry, we call it a distribution center. Basically, a location on a paved road so we can get the supplies to deliver. And then from there, hospitals and clinics could call us whenever they needed something and we'd send it. That was the whole idea. But we were wrong about how far we needed to fly, we were wrong how much we needed to carry, we were wrong about how much we needed to carry, we were wrong about how to think about cold chain, right? Blood needs to be refrigerated. We thought we had to do active refrigeration. We figured out, oh some simple insulation was all that's needed.

SAL DAHER: Yeah, that's all that it takes.


“Our first plane had a wingspan that was probably about four feet and what we operate now is 10 feet. So it's went from a plane you could carry under your arm with no problem to a thing that's a beast.”

KEENAN WYROBEK: It was sort of two and a half years from there to launch, but countless resets every six months basically. We'd make a bunch of progress, get a new system going in six months. By then we'd be like, "All right, so now we've learned a ton more from the customer, let's start over." Try to take as much forward from the component technologies we were making progress on but then rethink things. Our first plane was tiny. Our first plane had a wingspan that was probably about four feet and what we operate now is 10 feet. So it's went from a plane you could carry under your arm with no problem to a thing that's a beast.

SAL DAHER: You've got to put it on a truck. How did you fund this venture?

KEENAN WYROBEK: We're venture funded.


KEENAN WYROBEK: Yeah. Certainly in the earlier days we're venture funded and more recently we've been getting more and more support from NGOs.

SAL DAHER: So your venture backers, are they for-profit venture backers?


SAL DAHER: They believed in you enough that they thought this is something they could pay?


VCs Saw a Huge Opportunity

KEENAN WYROBEK: And what's exciting about this is that ... Well, obviously logistics itself is a huge market. But just taking this medical logistics market, just in the developing world, just in these remote areas outside the cities, is huge by itself. By any startup market standard. And what's exciting is that we can show them, hey, if we get 1% of this market in like five years, we're a great startup. If we get 5% ...

KEENAN WYROBEK: And you know, I think that combined with this obvious realization that logistics is gonna become more and more instant, more and more automated is sort of a general direction.

SAL DAHER: Yeah and it's a green field. Nobody else is addressing the problem. You find a good solution for this, you're way ahead of anybody else.


In the Midst of the Drone Craze in 2014 & 2015 Zipline Was Quietly Making Progress

KEENAN WYROBEK: Yeah. I think one other key thing that really helped get that credibility with those investors, which actually comes back to getting these customers first, so we started this February 2014. 2014 and 2015, a ton of drone companies were getting funded. It was just the thing that was happening. And we were super quiet about what we were doing. We had these customers, we were focused on it, we found a cattle ranch up in Northern California that we could fly all day over. And we were doing that. We were building and flying and building and flying since we started, as just recently, we've crossed over 10,000 flights at our test site.

KEENAN WYROBEK: And an investor wanted to check it out so we invited them over and they saw it. I think there was sort of this holy shit moment where he was like, "Okay, I visit a lot of drone companies and no one else is flying. No one else is actually doing it." Like he came by and we weren't demoing. At that point we were flying probably 20-30 flights a day there. And so all day long there was at least one or two of these zips in the air just drill delivering. And he was just watching it happen. And I think because we were so maniacally focused on what's the simplest way to make a customer happy, he was just like ... We made progress.

SAL DAHER: You weren't looking to be cutting edge of the technology, you were looking to be reliable. Sort of like a Jeep solution, not a Ferrari solution.


SAL DAHER: Yeah, very interesting. So where does your venture stand? How many flights are you doing a day? What's the growth and so on?

KEENAN WYROBEK: So we've been operating this first distribution center in Rwanda for a year. We started in October of 2016. We've flown over 2,000 deliveries since then. It's been going exponential because we've been controlling our growth by adding hospitals we deliver to carefully. We're at a point now where peak days are 30+ deliveries in a day in Rwanda.

KEENAN WYROBEK: But I should take a step back to give that some context. What we've been doing in parallel during that year is obviously learning a lot from that operation. And a lot of those lessons learned are addressed in software, in the control systems, how to fly through heavy weather, et cetera. And then a lot of other things we intentionally didn't do for that first system.



New Version of the Drone Is Built to Scale

KEENAN WYROBEK: We think of that first system like an alpha level system. You could go into the real world with it, but it was not scalable. Right? Not easy to port, not easy to manufacture, that kind of thing. So over the last year we've been designing the beta generation. It's really to scale. So designed for manufacturing, designed for field service, designed to make field operations much easier. Great example, the plane we have in Rwanda today, the alpha generation, you can turn that around in about 10 minutes. This next generation, you can turn around in about 60 seconds. So way easier. Maintaining it's much easier, way more modular and so on. Also makes it easier to build and all those kind of things.

KEENAN WYROBEK: So we are getting ready to field that. That next generation system is going out now. The other one would be hard pressed to make another distribution center worth of equipment. But this one, everything's ready to build lots of these drones and distribution centers, copies of the equipment. So really, that's what we're getting ready to do.

KEENAN WYROBEK: I've been talking about, obviously, the hardware piece. We've been getting the software ready so we can bring on delivery sites efficiently and things like that. And then we've also been getting our operations side. We sell service, we sell a service for a very concrete reason. Our customers don't want to buy drone technology. They want this problem solved.

SAL DAHER: They want to spend less money on perishable medical supplies.

KEENAN WYROBEK: Exactly. That's all they care about.

SAL DAHER: And what's the ROI for them there? Explain to me, give me an idea of the scale. How much they're saving by using your service?

KEENAN WYROBEK: Sure. It's a very complicated question to answer and honestly, an impossible one to answer. Especially when you get into the value of the health impacts and the lifesaving impacts. So as an example with blood, you know blood costs Rwanda between $100-$200 a unit by the time it's typed and tested and separated into components and so on. And so obviously those expiring on the shelf is very expensive. And so part of the way they look at this is they say, "We're eliminating that waste, which is huge." Not just straight up cost, but way more importantly is then every unit of blood that doesn't expire, actually saves someone's life.


KEENAN WYROBEK: Which is what I think of as the 10X value. And so our goal is to be cost competitive on the true cost value and then because of the immeasurable value it becomes obvious to do this. But yeah, so we sell it as a service so it's a recurring revenue thing where they pay us per month and that includes a certain number of deliveries and we go on from there.


Keenan Wyrobek’s Advice to Young People Interested in Drones

SAL DAHER: Think about what advice you'd give a young founder excited about drones, excited about robotics. What advice would you give knowing the experience you've had?

KEENAN WYROBEK: The advice is really simple. Go find somebody who has a problem that you're excited to solve. Actual people. People you can look at in the eye. You don't have to ask them what solution they need. But understand their problem well. And I truly believe everything that actually succeeds comes from that. I think there are lores, myths in the startup land. Steve Wozniak did not make a platform. He didn't. He made something that he and his friends wanted very much, the Apple Two motherboard. Or the Apple One originally. And he knew who he was making it for, he knew what they wanted to do with it, which was a certain level of experimentation. It's not that he knew exactly what they were going to do with it forever. But knowing that really specifically, he was able to create something really innovative.

SAL DAHER: And he had a store that would stock it.

KEENAN WYROBEK: Well yeah. Sure. That helps too. Well there's another thing. If you need a store to stock your product ...

SAL DAHER: If they had a keyboard.

KEENAN WYROBEK: Well exactly. But there's no better way to get someone to stock your product than if you can bring somebody who is a first customer with you. And say, "Look, no seriously, talk to this guy. He wants it." That makes everything easier. It makes fundraising easier. Going to a funder and being able to introduce them to customers, it makes product development, in my opinion, really easy, right? You get your engineers together. If you've got to get them close to that customer so you're not a middle man in the information there, and then you have this whole team now that can be working heads down to solve that problem for that person. And they won't go off into the weeds on some irrelevant tangent because they can literally talk to that person. That person can say that's not important to me. Or no, that is exciting. And it's very motivating for everybody involved.

SAL DAHER: That is wonderful. So is there anything that you'd like to say to our listeners that we haven't touched on?


Parting Words

KEENAN WYROBEK: Sure. So many of the medical breakthroughs that are accessible in the Western world, they don't make it to the people that need them the most. The billions of people who aren't here. Because of very practical reasons. At Zipline we're solving one of those practical reasons, but there are so many others ones. And I really think those barriers that are blocking those breakthroughs from getting to people who need them, those barriers represent very low hanging fruit for entrepreneurs. I've seen some incredible things done on almost no budget from a development point of view solving problems, creating maternal health kits, creating all these different things that development costs are low, you can literally have a product you're shipping in months. And the world is so small today. Getting out into the world and seeing this stuff first hand is not hard anymore. English, if you speak English, it's easy to travel the world now. All over the world. People will show you this stuff. They're hungry for solutions so get out into the world.

SAL DAHER: So the developing world is full of low-hanging fruit for a creative entrepreneur?


SAL DAHER: Excellent. Excellent. Kennan Wyrobek, I'm very grateful that you take out time from this extremely busy day, been a long day for you, to sit down with us here and to tell us about your really remarkable venture. Remarkable life, you had an impressive life before you tried this amazing Zipline venture. And I'm sure you're going to be doing many more exciting things. So thanks for sitting down with us.

KEENAN WYROBEK: It's my pleasure. It's great to speak with you.

SAL DAHER: Awesome. This is Angel Invest Boston. We're coming from Palm Springs, California at the TEDMED 2017 Conference. I'm Sal Daher.

SAL DAHER: I'm glad you were able to join us. Our engineer is Raul Rosa. Our theme was composed by John McCusick. Our graphic design is by Katharine Woodman-Maynard. Our host is coached by Grace Daher.