Just Gonna Send It
Just Gonna Send It is a podcast about the people behind the machines, the welders behind the hoods, the fabricators behind the sparks, and the stories that made them who they are. Hosted by Jim Belosic, CEO of SendCutSend, each episode focuses on the events that put our guests on the path of manufacturing, fabrication, and hands-on trades.
We don't cover much industry news or the latest tech. It’s more about why people build, how they got started, and what lit the fire. We talk about childhood influences and first jobs, side hustles and screwups. You'll hear conversations about what got people into making things, and the moment where they knew to "just send it" and make it happen.
Whether you’re already in the industry, or thinking about taking that first step, this podcast is a reminder that everyone starts somewhere, often without a plan.
Just Gonna Send It
Keenan Wyrobek (Co-founder/CTO, Zipline)
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Keenan Wyrobek, Co-founder and CTO of Zipline, joins Jim Belosic to talk about building one of the world's largest autonomous drone delivery companies. From creating ROS, the open source robotics platform that transformed robotics development, to launching lifesaving medical deliveries in Rwanda, Keenan shares the winding path that led to Zipline.
They discuss how Zipline went from an idea to operating around the world, why rapid prototyping and relentless testing are at the core of building reliable products, and how focusing on the real customer problem helped shape every engineering decision along the way. The conversation also covers robotics, startups, and the lessons Keenan has learned from turning ambitious ideas into reality.
One of my favorite vendors at the time was a a local bouncy castle company who was helping make prototypes of this airbag for us. Okay. And who was the first customer? Uh the public health system of Rwanda.
SPEAKER_02You're like, screw it, I'm gonna go do this to go to Africa and talk my way into people's cars or what?
SPEAKER_00Third grader engineering plus motorized suspenders equals pain. Uh but yeah, I just yeah.
SPEAKER_01Did you wear suspenders every day? No, I just thought you'd like a motorized suspenders. Were you watching a lot of like Linux or something?
SPEAKER_02Hey guys, welcome to the Just Gonna Send It podcast. I am your host, Jim Belosik. Today we're joined by a very special guest, Keenan Weirobeck, uh co-founder, CTO of Zipline. Yes? Is it CTO? Okay. Yep. Um for those of you Well, okay, everyone knows about Zipline.
SPEAKER_03I can't just in case they don't, though.
SPEAKER_02Okay, yeah. Sorry, we're joined by producer Dom, too. Um everyone knows what Zipline is, but Keenan, do you want to tell everybody what Zipline is in your own words?
SPEAKER_00Yeah, absolutely. So at you know, at our base, we're a drone delivery company. Um we are all about delivering healthcare when it's needed, the parts you need to do your job when it's needed outside of healthcare, delivering hot food and goods for people's personal lives. Uh, we do that in nine countries around the world. Um and uh yeah, including now here in the United States.
SPEAKER_02And so I I've known about you guys forever. Um when did you start? Like you said maybe 10 years ago or something like that?
SPEAKER_00We started we we started 12 years ago and started serving our first customer uh just about a little more than 10 years ago.
SPEAKER_02Okay, and who was the first customer?
SPEAKER_00Uh the public health system of Rwanda.
SPEAKER_02How do you land a customer like that? Like I guess maybe take us back a little bit, um, maybe a year or two before that first customer, that first delivery. What did that time frame look like?
SPEAKER_00Yeah, I I started, you know, I had the luxury of spending a couple years looking for my next big thing I wanted to work on. Before this, I started uh Ross, a robot operating system, and spent about uh decade getting that uh to become the the Linux of robotics, really. And um and I wanted to find something similarly impactful. At that time, uh Amazon was like drone delivery coming next year. And uh my wife's an epidemiologist, and I kept hearing these stories of public health campaigns, like vaccine campaigns that got stuck on logistics. So I was like, all right, let's I'm gonna go explore this. And so spent a bunch of time in Central America and Africa, just literally uh going deep with anybody in this world of logistics who would who would go deep with me. I literally cocked my way into the car of the head of the uh pharmacy system for a whole country and rode in his car for like six days as he was visiting all these uh clinics and pharmacies around the country.
SPEAKER_02What do you mean, talked your way into his car? You're just like hanging out outside of his car and you're like, give me a ride.
SPEAKER_00Basically, I was there, I was talking to someone who worked for him and I was explaining what I was doing, and I was like, is there any way I could just like go with you? He was like, I'm about to head out to go visit a bunch of sites. And I was like, how how could I possibly go along with you? And he gave me a funny look and they said, Come on. And I wasn't even packed for it, just went. It was awesome. Uh that that's not even the craziest story from back then of like talking my way into various like uh learning opportunities, let's call them. Um, but what's cool about that is that the the deeper what we went into it, it was really clear that the the supplies were there, the doctors were there, everybody knew what to do, but they just needed the supplies to do their job, which is something I can relate to, right? Like I've spent my whole career making hardware, right? If we make hardware, fundamentally your velocity and and your output's dictated by can you get the stuff you need to do your job, the literal physical stuff. And so I could really relate to these doctors, and and it was just like, okay, let's build a drone system that gets all these medical supplies, starting with blood, from point A to point B. And that's where we started.
SPEAKER_02Okay. That's so what were you doing? Like, did you have a job too? And while you're trying to get this thing off the ground, or are you just you're like, screw it, I'm gonna go do this and go to Africa and talk my way into people's cars, or what?
SPEAKER_00Yeah, I mean, I so at the time, so I was started Ross back at Stanford, spun that out into a company called Willow Garage that really got it going, uh, and then spun that out into a nonprofit to basically steward Ross. And so at that, I was unattached. I looked at a bunch of different potential startups at that point. I started some in ed tech and some other things that I thought were interesting, exploring various directions. Um, and so yeah, I was I was in a very luxurious place to just explore and not do any tech development, just really go understand the problems and to see if there's something I thought uh that I could solve uniquely and that like needed to be solved uh uniquely. And and I was literally exploring. So yeah, that was it was a very special time um of trying lots of things, meeting a lot of people, going deep on a bunch of problems that turned out to be dead ends, and then finding this one that I just really got excited about.
SPEAKER_02So you you said you had the the luxury to explore and do that. Yeah. Um luxuries are usually earned. So let's talk about the before times. Um so actually I'll go all the way back. Like as a kid, um, you know, are you are you messing with robots? Are you doing you know, are you taking stuff apart? What what's what are you like as a kid? What are your parents like in order to raise you, the the the man you've become today? And then we'll we'll end up at Willow and and more about Ross. But yeah, what are the the early evening? Oh, awesome parents.
SPEAKER_00Uh a teacher and a scientist willing to let indulge my curiosities like to two levels that I now realize were like very unique and a bit crazy. Um got to take apart absolutely everything, got to try to fix things, got to try to build things. Um yeah, that I think I still remember like one of the dumbest things I ever did in the third grade was like make motorized suspenders. Um third grader engineering plus motorized suspenders equals pain. Um but yeah, I just yeah.
SPEAKER_01Did you wear suspenders every day?
SPEAKER_00No, I just thought you had motorized suspenders.
SPEAKER_01Okay. Were you watching a lot of like Looney Tunes or something?
SPEAKER_00I I uh you know, I I I I I am willing to indulge now mentally, then physically, crazy ideas. Um and um yeah, built all kinds of crazy things. I still remember in the fifth grade making hooking all the desks in my classroom up by Morse code. And first of all, it was awesome that my teachers let me do this. I just thought it was a good idea. I can still remember like my classmates being like, who the the hell? A flashy light when I push a button here on other people's desks, like who wants this? Get out of here. Um yeah, but anyway, I thought it was a good idea. It wasn't.
SPEAKER_02I mean I think a lot of people, a lot of people have some of those opportunities, but the people who lean in, like the teachers that will let you do that, it's they'll let you do it because you're also like not a total screw up. Like you must have been getting okay grades and testing okay and stuff like that. Because I hear both sides. I hear, you know, oh, you know, I I get all this freedom, um, but then the teacher punishes me, you know, because I'm I'm not getting good grades. So you kind of have to have both, right? Is that is that possible? Can you can you get freedoms from teachers if you're performing? Or or is it just because the teacher was super cool and liked you?
SPEAKER_00Or honestly, at that age, it was the latter. The teacher was super cool and liked me and saw something in me, I think. Um grades were not amazing then. Grades got better later, once I kind of figured out how to like do good grades. That wasn't in the early days. Um mostly because I was mostly excited to just take stuff apart and build stuff. And you know, it was like it's like, no, no, learn to spell, which to be clear, I still haven't learned. Um they're like, well, you must learn to spell. I was like, who cares? Like I there's all this other stuff. And and uh yeah, just amazing teachers. I go, I was I was super lucky.
SPEAKER_02Okay, so so you're making motorized suspenders, uh, you're doing Morse code in fifth grade, uh, you you get through high school, um, you end up at Stanford. So what is what is college like for you? Did you know, like, were you like, man, I want to go to Stanford, or or were you just like, hey, I need to go to college because I want to do something, or did you just go to college and you didn't want to know what you want to do yet? Like, what was your path?
SPEAKER_00I so I I went I for college, I went to Johns Hopkins before Stanford was grad school. And I like I was I was really drawn to all the medical robotic stuff at at Hopkins.
SPEAKER_02Um are you thinking like in high school you're like you want to get into like medical cool stuff?
SPEAKER_00Or I was I was just like what are the pro I like I knew I like to build stuff, I knew I want to solve problems, I knew I love invention, and I love I just wanted to figure out like where the important stuff was. Um and um that was part of why I went to Hopkins. Uh in high school, I made a uh an underwater, an ROV, an underwater robot. Um that got me connected to a professor at Hopkins, uh who's you know, one of the world renowned um uh specialists in like the he built the system. This is Dr. Lewis Whitcomb, who built the system that went to the bottom of the Mary Islands trench and stuff like that. And so, like, you know, I got you know, I when I when I was visiting colleges, I visited there first and I just immediately connected with these professors. I was like, I have to go here. And um, yeah, what drew me to Hopkins actually was really awesome for me because Hopkins has a philosophy for grad students of only having uh three to five grad students in a lab, unlike Stanford, where you have like 15 to 20, which means the undergrads get to do a lot of grad student level like problem solving. And so that was great. I freshman year, I was like doing the crazy stuff I dreamed of doing um uh at in college. Um and just, yeah, anything I could do to work on um various things in medical robotics, got to work in the urology department there, um, worked on some underwater robotic stuff as well. Uh, but very quickly got the bug on the medical side and just went really deep in that space and built robots that could do like surgery on you inside an MRI field. So, like no metal, no wires, no motors. Um, that was stuff like that. Just where people were like, hey, if you could only use surgery while you're getting an MRI and you know, you could biopsy perfectly for these really rare cancers. And that was like, sweet, why don't we figure that? I got I want to work on that.
SPEAKER_02This this is really cool because like we go back and forth on, you know, college isn't for everyone. You know, sometimes sometimes people are better off going into a trade, and sometimes people say, Oh, I went to college, uh, but it was a waste. You know, I I didn't I didn't learn anything, I didn't get the opportunity. It sounds like you found some sort of magic there. Um, but it's it's because you were maybe you got like an insider track because you got to see you know that you talked to that professor before you even went there, you know. Is that totally part of it?
SPEAKER_00I think that's part of it. I think the insider track came from just like building from a really early young age, just like building, you know, just doing and like you know, I think you mentioned the college trade. We I I have a little team on the side at Zip Lime, we call the new products team that does all the really crazy stuff, uh that Zip on all in the early days. And I just lost my second long-term high school graduate intern to a startup to to be a founder of their own startup. The second time now in two in three years. Um, and uh, which is interesting because like I've now I now warned some parents when I find these really special high school grads who want to intern for this team. I'm like, hey, very good chance that they will not be going to uh to college if they come in turn here.
SPEAKER_02Straight into the pros.
SPEAKER_00Yeah. I think it's all I I for me, it was what I for me it's just all about rate of learning. I was it was I was having so much fun learning, getting to do all young old exposure from a really young age. And like that went all the way through through uh um college, grad school too, but I didn't finish grad school because I was just like the had this awesome opportunity to leave grad school and and I felt like just have a bigger impact on the world and learn more faster than I was learning in grad school and took it.
SPEAKER_03Were you doing like robotics clubs through high school and college for experience?
SPEAKER_00I that wasn't not in high school. Um we we did a whole bunch of like projects with other classmates that were just kind of weird and strange, uh, just because that's the friends that I love spending time with. It was like, oh, the school wants a starting system for the swim team. Like, can we build that? Like it's like we built one. Uh, was it awesome? Not as awesome as I imagined it would be. Or uh yeah, we built all we built various robot systems, like the ROV I mentioned. Also, just crazier ideas. The the we would just like sit around after school and be like, I wonder if you could like make a sound orbit your head. Um, sound like it's orbiting your head, and the pitch of the sound would be the distance like the like the space was around you. So you could like see in three, like at least a 2D plane. And we built that. You had to like hug a laptop, you know, not a laptop, a desktop computer on a cart with a battery behind you to make it work. Uh, and it barely worked for like 20 minutes before breaking, but it just ideas that hadn't that were terrible ideas, but really good for learning and just also for learning the process of exploring ideas. I still remember when one of my friends uh he was like, Oh, I'm gonna solve, you remember expo markers? He's like, I can recharge an expo marker. Uh like make a little thing you plug an expo marker in, and it worked. It worked beautifully. And I still remember when the school got a cease and assist from the expo company uh when this got some news. I was like, oh, that's interesting. What is that about? What's a cease and desist? Like I, you know, it's got all these interesting exposures to people like kind of thinking differently and trying things and um and and getting a little bit of exposure of like what is the real world outside of you know playing in your garage look like.
SPEAKER_02Yeah. Yeah, that's that's cool because like sometimes you know, we'll we'll encourage people to go the club route. You know, we're like, hey, join a club, it can it can pay off dividends, you know. If you have um a high school robotics club that can help you, you know, start uh on a on a better path towards college. And if you're in college and you join a club, then that can help you with a career, you know, after college or whatever. But it sounds like, or you can just wing it with a group of buddies and just do cool shit if there's a club or not, right? Absolutely.
SPEAKER_00And today is so fun. I meet so many people in elementary school, let alone high school, who who are, who can have access to technologies and learning on about these technologies that like I didn't get till I was in grad school. And I'm just so jealous because they can do all these amazing things. They just have an idea and they can try it and build it and learn from it. And I still think to this day, I I reflect more on my less on the crazy things I tried at that age that were total disasters to guide my thinking and problem solving to like succeed at Zipline today than like what I learned later in life when the pressure was on you really had to succeed. You just those early early opportunities to try things and fail and try things and fail and try things and maybe work and maybe work and and fail. Uh, you just I don't know. That that's where I got so many of my nuggets on how to think.
SPEAKER_02So that's that's actually something we should pull on a little bit. Um, we talk a lot about you know how people grew up, how were they they were raised. That kind of sets you up to be an adult. But if you're a listener and you didn't get raised like that, you didn't have those opportunities, it's like, well, what do I do now? I'm 35. You know, how do I um how do I get those opportunities now? So you're saying even if you didn't get it when you were little, you can still simulate that today by just trying stuff, just go out and do it, just experiment.
SPEAKER_00Yeah. Take take a problem, like it could be sitting in your house and be like, just build something for it. Just just start there. I think there's so many cool ideas come from like, you know, a problem you actually already understand and imagine imagining the solution to it. And and like even if it never works out, you're gonna learn a ton with everything you do. You're gonna learn new technologies, you're gonna learn like how to think about problem solving. Um, yeah, just just and it's so fun to learn now, right? You can AI can teach you, YouTube can teach you, you know, it's uh there's so much great open source out there. Um, you it's this the world's your oyster, and and so much has gotten so affordable um uh in terms of like technologies for prototyping and building. It's it's awesome.
SPEAKER_02So so you go, you're Johns Hopkins, you you're doing crazy no-metal robots inside of MRI machines. That's awesome. You go to Stanford. Tell us about Willow Garage.
SPEAKER_00Yeah.
SPEAKER_02I mean, that's like that's a big deal.
SPEAKER_00Yeah. This was the fun, that was a fun transition. So at Stanford, it's like, all right, I'm gonna get a PhD. I got all this funding for a PhD, I should do a PhD. So I talked to all these folks finishing their PhDs, being like, what advice would you give me? How'd your pro how did your journey go? That kind of thing. And um, you know, going into those conversations, like, all right, well, I I need to advance the field of robotics that, you know, some way, uh algorithms, math, whatever it's gonna be. Uh, but but instead I kept hearing the same thing in all these conversations with all these uh people getting their PhDs in robotics, which was like, yeah, my PhD journey was, you know, you know, years of clues in together a robot and collaging together the software, and then and then maybe you do some new stuff at the very end of all that and like get your video and then hope the robot doesn't break before the video's done, and then like get you know, graduate and get out of there, and and that robot goes in the trash heap. And that was like everybody's story over and over and over again. And um, it just gave me that itch of like, oh, there's an opportunity here. If there was code sharing, if there was open source and in robotics, um this the world of robotics would just be so much the world of uh robotics innovation would be so much more powerful so people could build on each other's results.
SPEAKER_02Because the PhD guys, the PhD guys are basically starting from scratch every single time, throwing it away. And so so you're saying, hey, what if we kept this like library of knowledge that you could repurpose? Okay.
SPEAKER_00Exactly. Everybody who could contribute contribute to it. And so we started that at Stanford and we got that going and tried to fundraise for it, uh, and eventually met the person who funded Willow Garage, Scott Hassan. And he's like, look, I'll I want to fund you. This is I'm you know, I want to I built, I made my fortune on open source, you know, in the in the dot-com boom. I want to pay it forward. I'm really passionate about robotics. You're clearly passionate about open source robotics, like I'll fund you, but you gotta leave Stanford. You can't do this at Stanford. I tried that back when I was in Stanford. You got to leave. Um come do it at Willow Garage. And we did. So we my my partner Eric Burr and I, we left Stanford, went to Willow Garage, got it going there. Eventually became all of what Willow Garage did. And yeah, it was just it went amazingly well. We built a community around it, we got a whole bunch of universities bought in and some companies in the early days. Um, and really the community just took off like wildfire once we got once we got it, once we got the base product good enough, we spent a lot of time with, we brought in all these grad students and said, okay, come come over for the summer and try this thing. Of course, in the early days it was useless garbage, right? They couldn't do anything. Uh, but eventually we we learned from them and iterated and iterated and iterated and got to the point where all of a sudden people were like, Oh wow, I can then just jump on this platform and and actually innovate on top of it. Um, and then like they all went back to their universities, and then now everybody uses Ross for robotics research and most companies. And so it was a pretty powerful uh thing that took off. Honestly, it was so funny because in the very early days of Stanford we had this deck, it said make the Linux of robotics, and it had the Rosie the robot on the front slide. And I remember the we pitched and got rejected by you know gazillion people for funding at Stanford um as like a donor type funding. And so many of their feedback was like, you know, you have no credibility, like this is crazy. You can't talk about the Linux of robotics. Stop, you know, don't be you know, you sound crazy saying this. Yeah, and it only took like three or four years before it just happened. Uh, and it was it was pretty wild.
SPEAKER_02That's that's incredible. So tell me, it you said that the guy was like, Oh, yeah, I made my fortune in open source, and and Ross is open source. How how do you make money in open source? I'm always unsure on that.
SPEAKER_00Well, he I mean he didn't make his funny money in open source, he made his money on open source. All right, open source. So he built companies on Linux and on the various open source software stacks for the web. And he's like, he wanted to pay for like the foundation of the future industry so people could build make their fortunes on you know open source and robotics. And that was one of the things that we wanted to do is is really enable innovation and and company uh you know, startups to not have to build all the little pieces just to get you know get going in robotics, but instead take a big common open source platform and build on top of it, just like you can on the web.
SPEAKER_02Yeah. So when you look at Ross today, like are you are you going, you know, are you proud of of where it is? Do you feel like you're watching from the sidelines? Are you do you are you still active? You know, what is what is that like to kind of watch it?
SPEAKER_00Oh, I'm not super active. The people who've led it since I moved on are doing a phenom much better job leading it than I did uh leading it. So yeah, super impressed with what they've done. Uh very proud of it, absolutely. Uh yeah, I think it just like when I think back on it, it's really just I I'm just so glad that you know my the the people who started Ross with me that we did ignore all the people who said we were crazy and just kind of really just believed and went for it. And um yeah, I was recently at a big Stanford, they created a new robotics center at Stanford and like touring the 50 or so robotics labs that are now at Stanford, and like everything is built on Ross there. And it's just crazy to see that just like it's like like what there's yeah, it's why I mean it's just it's also testament to open source. You build something really good, people are gonna use it, and that's just awesome and satisfying.
SPEAKER_02It's I I'm jealous that you have such a good legacy. Like I I think about legacy a lot where uh you know, if if I I guess maybe we have the same feeling, you know, we can walk around and aim, you know, or point to to some individual customers that uh their business exists because uh we exist. You know, so um We're kind of like the open source platform for sheet metal, I guess. Um, but a lot of people can can build on top of that dumb sheet metal and then make a real business out of it. So I'm pretty proud.
SPEAKER_00Your your legacy is way broader than that. Like I like the the the you know, as as this country really like rebuilds its capabilities in manufacturing industrialization, like you know, you were one of the first companies, I think really the first company who really showed w how possible that is. Um everybody else I know of showed it in a much smaller scale. Um, and you really showed like, no, this can be big in this country, and we don't have to innovate.
SPEAKER_02If we just if we we just keep saying yes, and then it gets harder and harder and harder. But bigger and bigger. We're still showing up. Yeah, that's awesome. Um okay, so so Ross is going good. And you know, what what does your exit look like? Are you you're like, um, I'm I got it to the point where other people can run it better than me and I want to go solve a new problem, or what what did that transition look like?
SPEAKER_00Exactly. Yeah, yeah. Some of the leaders who were part of that Ross world were like, hey, we want to go create a foundation to steward this long term, and they had a great vision for it. And at that point, like they were having a bigger leadership impact on the vision of it than I was, and they were also doing a phenomenal job of really building the community and the and the and that momentum. And um, and yeah, and I was I was getting that itch of like, okay, I've been doing it for maybe eight years at that point, and I was getting that itch of like, all right, I feel like I'm ready for the next like big challenge that needs solving. And that's uh so they went and did that, and I I went off and started exploring.
SPEAKER_02Can you can you talk about some of the other uh things that you explored that uh did not turn into zipline?
SPEAKER_00Oh, so many things. So many things. Um yeah, I I worked on an ed tech uh product in reading uh that was really interesting, but never never got traction. Um I worked on some other concepts actually with the co-founders that we did zipline with. You may have heard of some a product called a meeting owl. Uh that was one thing. Uh is that what we have? We have we have meeting owl, yeah. Excellent. That that was something we I worked on with my co-founders, and we were like, hey, we don't we're not excited about this, but one of the people we're working on was, and then he went off and did it. Um and uh looked at uh looked at things, um, some things in the industrial automation space uh that I think are only now becoming possible. And we looked at oh so many different things. Gosh, it'd be a long, it'd take us a while to take you through all of it. I spent better part of two years um just exploring a whole bunch of things. Some things like hands-on. Like I'm a big believer that if you want to exploring in theory is a waste of time. If you want to explore, you have to do, right? You can't like explore an ed tech reading app, you gotta build it. You can't explore like Zipline. We didn't go out to these governments and get you know this behind-the-scenes access to how these logistics systems are working by saying, We're thinking about building drones. We built a deck saying we are making drones to solve this problem and here's how it works. And even though it was all just sketches and nothing, no tech behind it, that's what got us in the door to get the learnings and the feedback we really needed. Um yeah, it was a fun two years of going deep and and you know, dead ends uh and some things we weren't excited about. Uh, but obviously, you know, this is something that we just got so excited about we couldn't help ourselves, and we went for it.
SPEAKER_02So, how do you know when you're exploring something, you know, you got two years and you're messing with a bunch of different things that are all over the map, right? Yeah. Ed tech to industrial automation to who knows what. Yeah. How do you, when you're exploring something, how do you know when to give up? When when you're like, uh, I went pretty deep, but I'm I'm bailing out. Is there is there common um common things that you look for or like a tingle on the back of your neck where you're like, oh god, I gotta get out of here?
SPEAKER_00Yeah. I mean so the exploration is is all is like you is a process of distilling um to the like the real first principles questions. Right? So for let's take zipline. It was like, okay, hey, in theory this makes sense, but why might it not make sense? Okay. Uh will these public health systems actually contract with a startup? Um, is there supply, right? If you being sitting there with drones to deliver supply, but the supply's not available, like what's the point? Um, is there gonna be demand, right? Are are are these doctors and and and these health systems gonna actually pick this up and change how they work to do this? And then and then you you kind of identify these like fundamental questions and then figure out, okay, what's the best way to answer them? Um and usually that's out in the field, really getting to know what's going on in the field. Um and I'll tell you like one story. Uh it's not a negative example, but a positive example was in a country that I shouldn't name because they asked us not to, where uh we were getting a tour of one of their big logistics warehouses for medical uh supplies, vaccines and blood and pharmaceuticals. And outside the warehouse, there's like a football field two stories high of pallets and uh boxes. And I'm like, Well, what's the deal with this stuff outside? And they don't really answer. And we go inside and get to learn about how their operation works. And later that night, I'm having dinner with one of the people who was giving us that tour. Um, and he's like, Look, I couldn't say this in front of everybody, but like all that stuff outside, that was all expired medicine. Um, like it's a big deal for us, it's very embarrassing, so we don't talk about it. But like it was one of those moments of like clicking of like, holy smokes, there's this applies there. So, like this one big question that was on our mind um felt uh that was one of what the data points that really kind of checked it off for us. And in other projects, we'd ask these fundamental questions and we'd get to a dead end. We're like, you know, I just don't believe, right? I don't believe this is really gonna work, or there this appetite's really there for this problem to be solved, or or that I could solve the problem. That's the other thing, too. It's like, okay, now that I'm standing the problem all the way, it's like, oh boy, this problem is like not actually something that I'm equipped to to go after um or ready to go after. So um is that is that does that answer your question?
SPEAKER_02No, that's amazing. Yeah, it's basically use first principles, or oftentimes we call it five whys. Well, just keep asking why, you know, until you get to the right problem. Keep asking that. And if if you can get all the way there and it still makes sense, then go for it. Like you'd be a good investor. Are you an investor? Are you an angel, Ever?
SPEAKER_00I've done a little bit now. I don't do much of it. I'm just so all in on zipline, but I I have done a little bit of some companies are just freaking amazing.
SPEAKER_02From what I've seen, I I think you would do okay. Uh I've I've had the opportunity to do a couple like very, very small angel checks. Um, or sometimes we trade metal for equity.
SPEAKER_03Yeah.
SPEAKER_02Every once in a while. But it's it's cool because I use the same approach. I'm like, okay, all right, you know, what is their likelihood of success? What kind of roadblocks are they gonna hit? Um assume that the market is there, then you know, what are their supply chain issues gonna be? And and sometimes even if we don't invest, it helps it helps the guy figure out like some of these questions, and then they'll come back in six months and be like, hey, I solved that that huge problem when you said no, I didn't say yes. So that's super cool. Um okay, so so you see the expired medicine, and you're like, All right, this is all starting to make sense. You make a deck that says, uh, we are making drones, they're just sketches, but we are doing this. Um someone just hand you a check and they're like, freaking go for it, and then you go for it, or or what did that look like?
SPEAKER_00Ooh, no, not not another the the I mean, so okay, how how the money followed the customers. So we went pretty deep in about five countries and then narrowed down to three and got to the point where we didn't have government contracts yet, but we had like we were advanced in the government contracting process in these countries. And I think that was our cheat code. Once we had that, then then investors were like, okay, this is this is real. Uh, that was also cheat code for the best talent. It's like you know, trying to get people to come do something crazy like this. It's like, no, no, we really have customers. Like, this is real. Um, and also the cheat code for, you know, I think it's also cheat code having those customers that we could spend this time with and be so deeply connected to what their real problems were was a cheat code for some of the hardest like leadership challenges of like what, you know, what are the requirements of this system? What can we cut? How simple can we make it, that kind of thing. Uh, having those customers gave us the clarity of uh uh to really, really focus. And not I'll give you one of the most intense examples of that. Um the we had just pivoted from what was going to be our first country, but we messed up the government contracting to Rwanda that became the our first country to launch. We were about three three months out from having to launch for the companies just to stay in existence. We were behind schedule, as you can imagine, right? Like no startup doing something like this is ever on schedule, I don't think. And um, and the Minister of Health starts emailing us pictures of people who were dying because they weren't getting blood. Um and you know, when you're doing a really complicated system uh technology, system of systems type thing, engineers can get lost in preference and style and like academic to type on arguments. But when you have that, it's like, no, no, no, what forget all that. Like what we we know what's needed that we know what the problem we're solving is is what what is the best way to go solve this problem for this real human, uh, and everything else that it gets cut instantly, right? The debates go out the window. Uh, and it's really not only motivating but clarifying. Because obviously the simplest possible system is gonna be the best system. Uh, and when you no one's ever done something like this before, it's really easy to make something that's like more complicated than it needs to be. But I found that again, having those customers who really needed it, um, it's so close so powerful for the talent, for the investors, and of course for the getting the architecture decisions right, what we could cut, you know, down to the very very essence of what we needed to ship.
SPEAKER_02That's that's powerful. Um, I'm I I I think every business needs that. Everyone might, I mean, you guys had incredibly powerful photos. Like, I guess, you know, like what's what's your powerful photo? What's your motivator? What's what's something that almost any company can grab onto to help them simplify and and cut scope and just you know latch on the MVP and get it out there. That's that's cool. I want to think about that. I think everyone should think about that. What's what's your photo of someone who needs blood? And I mean, like, look at Ukraine, you know, if you're getting shot at, you develop some uh some technology with whatever the hell you got and you get it out there. It doesn't have to be perfect, it just has to exist. So that's that's crazy. So this is uh like what's this time period? You know, you you go from concept and you know, going to customers because you don't have an actual product, right? You just you're saying, like, hey, I can deliver this using we're developing aggressively, very term. We so you go, they they get the contract, and then you're like, okay, cool. I got I got the contract. I have to fulfill the contract.
SPEAKER_00Yep.
SPEAKER_02You go to an investor then with your cheat code now, and you're like, hey, give me some money so I can hire some great people, start to build some stuff. How long does this take?
SPEAKER_00Like, what what yeah, that was awesome. The you know, Sequoia came in for us real early there, and and and like, you know, so life then, you know, we got access to this cattle ranch uh for testing, and and we decided we were gonna just all work there full time. So we got a little trailer, like we were at this cattle ranch. We moved in. Uh we had a little trailer, like it was just big enough. It was like, in hindsight, it was crazy what we tried to do in terms of manufacturing prototyping in this trailer on this cattle ranch. But it was phenomenal because we just flight test outside all day long. Yeah, I still remember when the partnership of Sequoia came out in the middle of a rainstorm, we had drones up in the air, and like there it was, it was I didn't realize at the time, but like they told us after they're like, look, we had looked at a lot of drone companies back then, drones were pretty hot back then, and there was nothing more clarifying than like that for them. Everybody else was like, I mean, demo, oh, but only in the nicest weather, and only in the and only and only and like and for us, it wasn't that we weren't demoing, we were just testing all day, every day, drones in the air. And and like, of course, it was testing in the rain because it rains all the time where we're in Rwanda, and like we knew we had to fly in the rain, so we were never gonna stop flying in the rain in our test site. And um, yeah, it it was it was a it was a it was a grind. It was we were figuring out the product and the requirements as we were going, of course. Um and you know, this was a fixed wing drone. So picture a you know 10-foot wingspan, sort of large RC plane looking um drone. And uh, we did fixed wing because again, we're so deep with our customers, they cared about two things range and low cost. And nothing, almost nothing else mattered. And if you're gonna do range, you're gonna do fixed wing. You're not gonna do a vertical takeoff and landing. That's like as a 10x advantage, it's not a subtle trait. Um, and so we knew that. But then of course now you just like how do you deliver? Because you can't like land to deliver. How are you gonna take off and and uh and and land um in all this crazy weather? And so we really focused on like how are we gonna do that? And turning through, you wouldn't believe the stuff we prototyped about how to land and how to launch. Oh I I I uh yeah that's what we want to hear about.
SPEAKER_02I mean, talk about the prototype process. What what what were some of the early ones? What were the biggest failures? Um, what were the biggest learnings?
SPEAKER_00Yeah, uh, I now understand the physics of a um of a stunt airbag very well, because I was very convinced that a stunt airbag with a drone belly flopping onto it was how you should land at steel.
SPEAKER_02The stunt guys, when they jump off of a building, they land, they they land on their back in the big airbag thing.
SPEAKER_00Exactly. Yeah.
SPEAKER_02Okay, so a fixed wing, you thought a fixed wing was gonna like flip on it.
SPEAKER_00Like, that's how we should watch the service. It turns out that the impact energy just of the of the material of an of a bag. Um yeah, I I remember one of our one of my favorite vendors at the time was a a local bouncy castle company who was helping make prototypes of this airbag for us. Uh and uh they're like they they uh the I I swear they coasted me the first time because they were just like this guy, you know, like what this order just came in with this weird sketch on literal pencil sketch of what they wanted. Can't be real. And I was like, I went and knocked on their door. I was like, no, no, seriously, please can you make this? Um and uh that was one of the crazier ideas, partly because uh on a weird it was held to the ground with uh eight-foot uh stakes, you know, the ones using like circus tents. Um yeah, and the big a big wind came in one day at our test site and just ripped like 38-foot stakes out of the ground and went flying across. Uh that was one that was one of the many moments where we were like, oh geez. Uh there's certain things that we just didn't like, you know, as we were prototyping fast, we didn't quite see coming. But yeah, we tried so many ways to land that drone in all weather. And uh we actually started operating in Rwanda with a system that used off-the-shelf deep sea fishing reels and poles and a line between the two, where the drone would come in, catch that line between two fishing poles, and the fishing reels would go and then it would land on a bouncy castle mat.
SPEAKER_02And that's how we started something like if you're landing on an aircraft carrier, like the tail of a pick grab it, and then exactly. Yeah, that seemed that seems pretty good. I guess you gotta go wind all that stuff up after you're done.
SPEAKER_00It was operationally nightmare, but like we could do it, and we start operating. Now that the system is like fully computer controlled and off the ground and resets itself and it's beautiful and done well, but gotta start somewhere.
SPEAKER_02You guys were using nets at one point, right? It wouldn't it go into a net? We tried nets. Okay, yeah.
SPEAKER_00We tried nets. Uh the problem with a net and a drone is it's very hard to the net damages a drone. Yeah, the reason actually I'm talking about landing so much is like in aircraft in general, all aircraft, hard landings determine the life of an aircraft. So if you reliably gently land that plane, planes can almost last forever. Um, because the lows when you're in the air and the sort of you can design and test for that very well, but it's the hard landing that gets you. And every drone we'd ever met, everybody we met in drone space was telling us stories about how, like, yep, we thought we had a drone that would last for 10,000 cycles, only lasted for 100 because we didn't understand hard landings. And so it's like, all right, we have to have a reliably gentle landing solution. And so we were on a quest to figure that out. Um, and we ended up with this aircraft carrier system because it was still to this day, far and away the most reliably gentle way to do it. And uh yeah, we've done it now. That that system has now landed over two million uh flights in commercial locations.
SPEAKER_02That's that's amazing. How how many uh like hobby grade RC planes did you guys go through? Like, did you are you are you do you start like when you're prototyping, do you sometimes go to like off-the-shelf stuff just to understand the basics? Yeah.
SPEAKER_00Yeah. I mean, the way when we in those early days, we just we break the problem into like we think are the like big hard parts of it, whether it's a product question or a technical question or whatever it is. And so in those early days, there was one team just working on the avionics, just the electronics and software to reliably fly in crazy weather. And they were doing all the testing on our off-the-shelf uh RC planes. And then a separate team was like, all right, what's the actual aircraft that can carry the payload we need? And they were just working on that with you know real control of controlling of those prototypes, and totally separately off on the side, we had um we were developing the landing system uh and we tested that. We took a launcher, like a pneumatic launcher, and a big old plywood drone, literally plywood wings and everything, same weight and same, and we'd uh we would fire the launcher right after the landing system to test the landing system. And and these we could all work uh independently, uh, each each each effort iterating at its max velocity. And so those are the early days, and and it was so fun back then for Zipline because when I say each team, we're talking like one or two people per team. Uh, we were only a dozen people, the whole company, and uh and really trying to figure out the first system that could serve customers.
SPEAKER_02So you said the early days were very fun. I mean, what let's skip forward to today. Like what what are the challenges today? I mean, as the company's evolved over a decade, um, I mean, you guys are solving real problems still, you have real customers, way more customers. Um is it still fun? Or, you know, what's I I I know they're still fun, but also like what's the grind that you didn't expect?
SPEAKER_00Oh, the grind. Yeah, this is definitely one of those stories where we didn't do this because we thought it would be easy. You know, I'm sorry, we didn't do this because it was easy because we did it because we thought it would be easy for sure. And everything has been weather is a good example. Maybe a hundred challenges over 10 years around weather. You know, all our use cases need all weather operation. That's just been, you know, and so um, you know, good example today with our second product, which is our precision delivery platform that delivers in a metro area, where the the fixed wing platform we were talking about, that's for going from a metro and serving like countryside. Um, this precision delivery platform, um we have uh we have a weather chasing team. They have a trailer with four docks on it, uh, four drones. And yeah, right now they're in uh Panhandle, uh, Florida, uh, because that's where we think there's gonna be the most uh probability of encountering lightning strikes. Uh before that, they were at this uh our weather team found a uh the windiest ranch in the United States, and we rented access to this ranch and flew there until we got good at it. And before that, we were up in upstate New York um flying in icing conditions, um, chasing the icing conditions up there. And um, it's just the weather has been all about understanding it because it in a lot of the stuff we do, no one else does it, right? Or if they do it, it doesn't apply. A good example, I just mentioned icing conditions, right? In a commercial airliner, right, that flies in these conditions, they actually ducked their exhaust heat into the leading edge of their wing and vaporize the water in icing conditions, right? It it works, it's it does not translate to a small drone. Uh we don't have the equivalent of like exhaust heat from a jet engine to duck it in front of our exactly um power there is this mind-boggling. Um, and so we've had to come up with our own solutions to these problems. And and icing is this classic example in engineering where it's it is extraordinarily complicated. There's like dozens of types of ice, and it's the way that bonds to materials varies and the way it accumulates varies. And and so we've spent a lot of time really making sure we understand these problems well so that we know we're uh in-house engineering and testing um the way we should. You know, maybe another way to answer your questions, I this is something that like you probably don't appreciate from the outside of Zipline, but half of all what the work we do at Zipline, like, and this is at every level, is is test. Um, it's not new development. It is developing you know hardware loop testing systems. It's like the weather chasing I mentioned. It's our flight high-volume test sites. Uh, we do phenomenal amounts of testing on the ground. We have a whole building now dedicated to testing, uh ground testing, running 24-7. Uh, and we're getting really good at basically how to break things in representative ways so we can speed up our development process and reach ever higher levels of reliability. And um, and that like and basically half of all of our work across uh like the history of Zipline has just been on that side of the equation.
SPEAKER_02So you're yeah, you're not developing new vehicles or anything, you're just making the existing vehicles hardened, you know, against all kinds of weird stuff.
SPEAKER_00Exactly. And and figuring out how and not just harden them, but like get better and better and faster at testing them. So if we like on our our existing plot um products, we you know we want to make them better by improving something about the hardware. It's all about how quickly can you re-validate that like that that improvement strictly made the system more reliable, didn't have aggression, and somehow uh how quickly can you do that is uh fundamental to our velocity of the company and our velocity of getting these solutions out to our partners. And so by basically hiring people who are really creative and really good at breaking shit uh uh in all kinds of amazing ways, like we've built this amazing capability that enables us to take new things and ship them uh, you know, well, by aerospace standards incredibly fast.
SPEAKER_02Yeah, we always talk about how we can test as much as we want, um, but our customers are endlessly. Creative and they will figure out new ways to break things. What are some of the weirdest like edge cases you've had where you know you test and test and train and prepare, and then you release it out in the wild and you're like, well, we weren't expecting that.
SPEAKER_00Yeah. Oh, there's so many good stories. Um, so a good hardware technical one that uh is around electrostatic discharge. So when we started Zipline, we were flying our drones, and I knew about a thing called triboelectric uh charge uh electrostatic discharge buildup. And that's basically where it's what happens when you take a balloon, rub it on your head, and then your hair stands up. Same thing happens on a on an aircraft. Air flying over the aircraft builds up static electricity, and then eventually that static electricity will use to discharge or equalize through the aircraft. If you have a metal plane, which we don't, then it sort of equalizes through the skin. And if you don't, it'll find a path to equalize. And so we were testing for this. Um, but we were we were flying a bunch and we were finding um every time we would fly through tur this special sort of shear layer turbulence, uh, the two nodes on our wing we have for redundancy were going out at the same time. And we're like, oh, we these are so isolated, this can't be happening. Must have to be to do with this turbulence and the mechanical vibration. So we were going deep and deep and deep on that. Uh, and after um many, many sleepless nights of trying to figure out what was wrong with this problem, uh, we finally got someone on the phone who was like, you know, who knew just was an expert in uh electrostatic discharge in flight. And he's like, you know, there's another type of ESD in flight. No, what, what? And he's like, Yeah, when you you know, when you go through these shear layers, the things on a like when on a commercial airplane, they have a big jump on the plane. You go through those shear layers, the electrical potential of the air on either side can be as much as 50,000 volts different from difference from one side of the shear layer to the other. And we're like, oh, so you we get these big turbulents. So we thought it was mechanically hurting us somehow, we didn't know. No, no, what was actually happening was like hitting the whole drone with an ESD gun at the same time. Um, and that was equalizing like through this common path that was like way out of the wingtips and knocking out these two nodes the exact same time. And uh yeah, that's a fun hardware one that was like one of those things where like, you know, we had a bunch of smart people who were thought we were fairly knowledgeable and we were going down and reading, and it was just like an area of knowledge we knew nothing about.
SPEAKER_02And if we hadn't been making those phone calls to learn, we wouldn't have we we call it old man knowledge, you know, just trying so hard at a problem, and then you call your dad and he's like, No, just turn it upside down. You're like, Oh damn it. You just had no idea. With like the the it seems to me like the pace of like robotics and AI and everything is scaling so crazy fast, like 10 years ago it was moving fast, and then five years ago it was triple the speed, and it seems like just in the last year, everything is is like exponential. What what excites you the most about like what the future is? We've talked a lot about the past of Zipline, we've talked a little bit about today's zipline. What are you excited about? What's what's something that you've never been able to do and you might be able to do very soon?
SPEAKER_00What has me excited is less about what I can do and more just like it is becoming easier and easier to make take robotics technology and make it useful. And and the like I recently met this this middle schooler who built automation technology for his family's factory, right? Like using using this technology, that's how accessible it is. That is so freaking cool. And I'm a huge believer in just like human capacity. And um, I've always subscribed to the the more Japanese sort of way of thinking about uh robotics, which and automation, which they like that that word just in that cultures means productivity. It's a productivity tool for humans. It's uh it helps humans do more and be more capable. And I'm just really excited about how democratized it is and how anybody can get their hands on it. It's so much easier to learn, it's so much more capable, it's so much easier to do much harder robotics uh applications than it used to be. Um, and uh yeah, just it I are all around the world. I meet people who are just like, they're just like, I wanted to do this, and like I want to. I I met this guy out in uh in uh in Tanzania who's like built a full-on coffee sorting system, just himself, like with like a couple hundred dollars worth of hardware he was able to got his hands on, and just like that's how his family's coffee like sorting went that they used to do manually in the over the kitchen table, and now they just pour the beans in and it just sorts itself and it's just so cool. And I think you know, I think the the future of what humans are gonna do now that uh anybody could get their hands on this, and if they have an idea or problem they want to solve, they can do it now. Uh it's so cool.
SPEAKER_02Um democratized is is a good way to put it. It's it's accessible now. Um it's accessible and and and easy. I I think I think in the past it was accessible, but you had to go read a bunch of books, you had to take classes, you had to to do some sort of apprenticeship or something, and now you can do it in a weekend. It's pretty amazing. It's so cool. So Keenan, uh we we end every episode with the same question. Um what's one piece of advice that you'd give to someone who's listening uh who's hoping to build something on their own?
SPEAKER_00Yeah. Oh, good question. Um if you're working on something that is complicated, uh make sure you understand what matters most about the problem and then come up with a simple solution to that part of the problem. And then design the solution to the rest of the problem. If you do that, you'll end up with a robust, simple overall system. I wish I had learned that lesson earlier in my career. Uh yeah, start start with the simplest possible solution to the hardest, most uh important part of the problem. I'll give one example. So, like our first uh that fixed wing drone system, our first long the platform, we call it the long range platform that I was talking about, we launched our business on. Like the fact that we understood so deeply that what mattered most about that problem was long range and low cost, it gave us conviction to do fixed wing, which many people were like, you got to do VTOL, the industry is going VTOL drones are the future of drones or VTOL. What are you doing with this fixed wing drone? You're like in the past. It gave us conviction to like ignore all those people and do the thing that actually made the yeah made financial sense and made the business work was being understanding that was what mattered most, and then coming up with the simplest solution to that long range low cost, which is fixed wing. Um, and then again, flowing the solution to the rest of the problem around it. Like, how do you launch it? How do you land it? How do you deliver? That all came second. Um, yeah, if we hadn't sort of followed that mental model of approaching the problem, uh Zipline wouldn't be around today.
SPEAKER_02Yeah. Dude, that's amazing. This has been super incredible. Thank you so much for taking the time and being on here. Um, if anyone wants to learn more about you or more about Zipline, how can they find you? What's what's the best place to learn?
SPEAKER_00Yeah, absolutely. Check us out, zipline.com. And uh yeah, you can you can find me on X.
SPEAKER_02Cool. All right. Uh, thanks again for your time, and we will talk soon. Cool. Thank you so much. This is a great conversation.
SPEAKER_03Thank you.
SPEAKER_01Do we have an outro? Am I supposed to say something?
SPEAKER_03Can you say like thanks for listening? Love to have your feedback. Find us uh on Insta and YouTube at Suncutsen.
SPEAKER_02Can we just use that?
SPEAKER_03Thanks for joining us on today's episode. If you have any questions, comments, or concerns, you can email us at just gonna send it at sankcutsen.com or find us on socials at sank cutsen at gymbelosic. Don't forget to subscribe, download, and share with your friends. Thanks again for listening. We'll catch you on the next episode.