[[Home|🏠]] <span style="color: LightSlateGray">></span> [[Interviews]] <span style="color: LightSlateGray">></span> August 8 2022 **Insider**: [[Peter Beck]] **Source**: [SmallSat 2022 Keynote](https://www.youtube.com/watch?v=0d2VZezle4s) **Date**: August 8 2022  🔗 Backup Link: https://www.youtube.com/watch?v=0d2VZezle4s ## 🎙️ Transcript >[!hint] Transcript may contain errors or inaccuracies. Hi everybody. Firstly, my sincere apologies for not being there in person. I got as far as the airport lounge and it looks like I caught some kind of food poisoning. As I was standing there at the gate about to climb on an aircraft, profusely sweating and shaking and a tummy tied in knots, I decided that was going to go one of two ways—either very well or very badly—and I thought the better of climbing on the aircraft. For all people involved, I'm confident that was the right decision. So I'm incredibly sorry that I can't be there in person and we have to make do with this today, but I also want to thank all the organizers at SmallSat. It's a great honor to do the keynote, and thank you for being able to find a way that I can still do it from the other side of the planet. ### Early Beginnings My talk today is titled "From Drainpipe to the Moon," which really summarizes Rocket Lab's journey to date. A little bit about me—I was born at the very bottom of the South Island in New Zealand in a town called Invercargill. That town makes Logan look like New York, it's so small. One of the very first memories I have as a child is actually standing outside with my father, and he pointed to the sky and told me all the stars in the sky are suns, and those stars have planets around them. From that point on, I was hooked. I was going to do something in space. Wasn't sure at that point what it was going to be, but definitely going to do something in space. The challenge, of course, is that down in New Zealand there was no space industry whatsoever. So anything I was going to do was going to be a little bit of a different path. I can remember my parents got called into my high school because the careers counselor advocated that my aspirations for what I wanted to do with my career were thoroughly unreasonable and that I should just go down to the local aluminum smelter and be a fitter-turner because I was very good with my hands. Luckily, I had parents that were really supportive, and I was able to follow my dreams, but it's been a crazy road and it continues to be a crazy road—and a fun one. ### Early Funding and the Mock-Up Rocket I'm going to skip a whole bunch of history and just bring us right up to the point where we raised the first Series A round from Silicon Valley. Those were completely different times. It looks like we're heading into some economic headwinds, but up until then, the market's been incredibly frothy. When I raised the first amount of funding for Rocket Lab, it was a five million dollar Series A. For that five million dollars we built a composite tank so we could prove that we could build a composite liquid oxygen tank, and our very first electric-pumped rocket engine. Up until relatively recently, raising 50 million, 60 million, or even hundreds of millions was pretty standard. I can remember running around Silicon Valley with a one-tenth scale model and an electric pump under my arm through the various VCs, trying to convince them that this guy from New Zealand was going to build this rocket. It all seemed very fantastical at that time. ### Ignorance is Bliss: The Drain Pipe Mockup This slide is really entitled "Ignorance is Bliss," and this is where the drain pipe comes into it. I've always believed that to actually take on a big project, it's great to physically inspire people with things that they look like they're going to build. The very first thing we did with our newly minted five million dollars was build a mock-up of the rocket. Of course, the only piece of pipe that we could find that was a similar diameter was a piece of drain pipe. So that's me there, considerably thinner—can't even fit into that suit now—standing beside a mock-up drain pipe of Electron. I think the saying is that no good plan survives first contact with reality, and that was certainly the case for the early days of Electron. There were lots of new technologies that we had to develop, lots of new systems, new launch sites—you name it. It was a tremendously steep learning curve. But where we ended obviously is a completely different place. We took that drain pipe and developed it into a very reliable and resourceful launch vehicle. ### The Challenge of Scale I think Elon once famously said that building your first rocket is easy—the hard part is building them over and over again. I can testify that is absolutely true. The first flight of the launch vehicle is really hard, but you've got a year or so to do it with a whole bunch of engineers pouring over every detail. The real challenge comes when you're at about flight 10. I reckon it's 100 times harder to build your tenth rocket, and by flight 20, it's like a thousand times harder to build your twentieth rocket. I think that's been consistently said by others within the industry. The big difference is that the first rocket, everyone's pouring over in great detail. But by the time you get to your 20th rocket, it's all technicians, work instructions, ERP systems, MRP systems, supply chain, factory tracking, quality tickets, and all those kinds of things. It's immensely more difficult to do something 20 times over and over again reliably than to do it once or twice. That was my biggest learning. In the previous slide, "Ignorance is Bliss," when you start out, I can remember all of these things were going to be super easy, and it was going to be straight sailing. I didn't understand why somebody hadn't hurry up and done this in the past because it seemed so incredibly straightforward. But boy, reality is the great leveler. ### Building the Launch Site Not only did we build a new rocket, we actually built a launch site as well. I think a lot of people think that we have a launch site down in New Zealand because I'm a Kiwi and that's where we wanted to build one, but that's not the reason at all. We built a launch site down in New Zealand because if you fast forward to today or even another five or ten years, it's very obvious that the more traditional launch sites become busier and busier. When you're trying to build what ultimately will become a high cadence launch system, that becomes very challenging. When we set out to solve this problem, we didn't just stop at the rocket. We looked for a way that we could launch really frequently and also have ultimate flexibility. With a small rocket, you're a dedicated service, so you go when the customer is ready. When the customer is ready and when you're ready, the other variable you don't want to have to cross is "is the launch site ready?" So we're able to decouple and solve that problem by just having our own launch site. If you're looking to build your own launch site, then you want ultimate flexibility. You need a site that can shoot a large array of inclinations and trajectories. You can't launch over land, and you kind of resolve yourself into this place where boy, wouldn't a small island nation in the middle of nowhere be an ideal launch site? So that was the rationale and the reasoning behind actually having any operations in New Zealand—it's this launch site. Building a launch site isn't just rocking up to somebody and a launch site appears. It takes a lot of work. We went up and down not just the country but the world looking for a launch site that would give us everything that we wanted. We found the Mahia Peninsula. If you trace New Zealand and look down the side of the east coast, you'll see a bit jutting out, and that bit is the Mahia Peninsula. We identified the ideal piece of land and then we went and saw the landowners and farmers. Lucky for me, the farmer was looking to diversify his business. It's actually owned by a local Iwi trust, and they were looking to diversify. I think they were probably thinking maybe different kinds of cattle or different kinds of sheep, or maybe if they want to be really adventurous, some deer. I don't think anybody was looking to diversify into launch vehicles, but ultimately we were able to strike a super great relationship and we started to build the launch site. I don't think everybody realizes or appreciates, unless you've been there, just how remote this launch site is. It is literally in the middle of nowhere, which is great for a launch site—that's exactly what you want—but it makes it very challenging to get to. It used to take us a couple of hours to get from the last turnoff of the road right down to the launch site. What is now a 45-minute journey required us to build 32 kilometers of road. We had to put power and infrastructure and comms—a tremendous amount of just bare bones infrastructure for the site to be able to begin in its first instance. ### The Launch Complex Today This is the launch site as it stands right now. It has two completely independent launch pads. Both these launch pads operate absolutely independently of each other, so you can be fueling on one and wet dressing on the other, or launching on one and preparing on the other. We have three clean rooms on site with full customer facilities. One clean room is hydrazine compatible. We have a tremendous amount of infrastructure there. We have a hangar that's capable of holding two rockets at any one time. We pretty much keep two rockets down at the launch site these days. As we're processing on one pad, we're preparing on the next. What ultimately ended up here is a launch site that's capable of launching every 72 hours. We're licensed to launch every 72 hours, but probably the biggest learning from all of this was our assumption that we're going to have a dedicated service, so flexibility is going to be needed—it really pays off. If we poke our head out one morning and don't like the look of the weather, we'll just roll it back in and not launch that day. Likewise, if a customer has an issue with their spacecraft and they need an extra day or two to resolve, no big deal—we'll just roll out when the customer is ready. That becomes very challenging at traditional launch sites where you're scheduled and you have very short time frames to get your launch away. That vision of what we wanted to create there has been super important and really valuable. ### Expanding to Wallops Island Now we have a total of three pads in two hemispheres. We built a pad at Wallops Island, and NASA assures me that the AFTS system will be complete this year for a December launch. We're really excited about that. In fact, this month the launch vehicle will be shipping there. We've already had a launch vehicle there and wet dressed it and it's ready to fly. We put that launch vehicle back into circulation and shipped a new one there. We're all hanging out for this launch out of Wallops at the end of the year. That launch site was really designed around one express purpose, and that was responsive access to space. What everybody sees there is a launch pad. What you don't see is a whole ICBF facility that's just back from that where we're able to process three Electrons. We have clean rooms, hazardous fueling operations and facilities there, skiffs, everything you name it. That's designed for rapid call-up. Responsive space is obviously a big topic. It's something that we're very passionate about, and we built this launch site for that express purpose where right from call-up to the pad to launch is capable of doing it in what you would generally measure in hours instead of days. That launch site—we're really excited to fly out of the first time and provide that national capability for all of our customers. ### The Rutherford Engine No good rocket is any good unless you've got a good engine. We named the Rutherford engine after Ernest Rutherford, and Ernest Rutherford was an incredible physicist. In fact, all of our engines are named after physicists. He had a saying that I really loved, and it's kind of embodied the whole Electron program. That saying was, "We have no money, so we have to think." You have to understand that the development of the Electron program, compared to rocket development programs or privately funded rocket development programs of today—we did that whole thing for tens of millions of dollars. That whole program wasn't hundreds of millions, which has sort of become the standard today. We really built Electron and everything on the smell of an oily rag. The engine itself is an electric-pumped, additively manufactured engine. We 3D print out of Inconel the vast majority of the engine. I can remember that picture is actually me announcing that engine at Space Symposium—I think it was about 2015. There were obviously a lot of raised eyebrows about that engine when we first announced it. I think everybody thought that 3D printing a rocket engine was a little bit silly, and certainly an electric pump cycle was a bit silly. But as it turns out now, 3D printing a rocket engine is just the standard way you do it. When we started mucking around with that technology, I remember going to an additive manufacturing show. There were kind of two things that came out of it: they're great for printing bottle openers that you can take away with you from the show, and a metallic 3D printer was also great at printing a cat's prosthetic—that was the application of note during those times. Nobody really contemplated trying to push it to building an entire rocket engine where you have a throat that's running red hot or white hot sometimes, and really highly stressed engine components like pumps and injectors. So we really pioneered a whole lot of technology there. We did a lot of material science and development and printing development. It was a big bet. But where we are now—every 24 hours a Rutherford engine comes off a printer, and it's an incredibly reliable, high-performance, mass-produced product at this point. Then the electric pump cycle, departing from your traditional gas generators or other cycles, was once again a fairly big risk. But with a small launch vehicle, it's all about residuals. Mass fractions are really tough, but it's all about residuals because when you have a payload of 100, 200, 300 kilograms, if you've got 50 liters or 100 liters of propellant left in your second stage or even in your first stage, you know you almost lift no payload. The great thing, and really the reason why we went down an electric pump route, was that throughout Electron there's a continuous level sensor in both tanks. It's continually closed-loop optimizing the oxygen-fuel ratio to ensure absolute depletion. In a first stage, the other advantage of an electric pump is you can load sense off the pump and actually shut the engine down when you're truly out of propellants—not at some arbitrary level sensor point, but you can actually suck the tunnel dry. In the first stage, when we ascended first stage, we actually sucked that entire stage, both propellants, completely dry. We can actually tell when the LOX is in the tunnel and basically judge the amount of remaining LOX in the tunnel, let alone in the bottom of the tank. The second stage is even more important. There are a number of closed-loop control events that occur during the second stage ascent where, once again, we're continually optimizing the oxygen-fuel ratio to ensure we have equal residuals and don't run out of one versus the other. That's really the magic of the electric pump—you can load sense and you have such fine control over residuals. The engines now—we've put 290 of them in space, and it's a very mature engine, very high performance. 346 seconds ISP in vac in the upper stage engine. For those who build little engines, it's ridiculously difficult to get those really high ISPs out of a small engine. ### Carbon Composite Structures The other big bet we made was on composite tanks. At that point, nobody had flown a rocket to orbit with a carbon composite tank, so this was a new technology that we really pushed hard. An Electron first stage and second stage tank has got a 1.8 millimeter wall—sorry for the metric as opposed to imperial—but a very thin tank wall. Composites get a tough time about being really expensive and costly to produce, but the reality is when you only have a 1.8 millimeter wall, there's just not that much material in there. We've developed and pioneered some really great mass manufacturing techniques. A first stage tube is literally tape wound and then goes up into this machine you can see in this image here called "Rosie," where all of the machining process is done to it. Then you slide a couple of bulkheads and tubes and just glue it all together—so very low cost in the end and super high-performance structures. Electron is around a 90% mass fraction complete vehicle, so really high mass fraction, and that's really enabled by all the composite structures that are throughout the launch vehicle. ### First Launch and Failure Once we had our engine sorted and our tank sorted, it was time to put one on the pad. We closed the first investment round at the end of 2013, so we really started work on Electron in 2014, and we had one on the pad by 2017. And of course, we had to build the pad as well. [Referencing video] A lot of you probably have seen it, but the thing to really watch out for here is at the end of the video you'll see some antennas on the ground kind of searching around in the sky. It was a perfect flight—perfect liftoff, perfect first stage burn, and great separation, second stage burn, fairing separation. We were just sort of trucking on home to orbit. 90 seconds—I think it's about 90 seconds I've ever want to hear our vehicle terminate. Of course, we're all standing there thinking, "Well, it's right down the IRPs, it's bang on trajectory, why would we terminate?" This is kind of a tough lesson that we all learn in the launch vehicle industry. Those dishes on the ground with the flight termination dishes—you may have seen them kind of wandering around looking for a signal. Well, the reality is there was one tick box in a piece of software that wasn't ticked for error checking, and as the errors accumulated, it continued to search around for the rocket and couldn't find it, and then ultimately was terminated. I actually took a screenshot of that software page in that tick box and framed it, and it's sitting in the entry of Rocket Lab as a reminder to everybody just how tiny the error can be before you lose the lot. But otherwise, what would have been a completely successful flight, we presumed to orbit—the vehicle was completely nominal. ### Current Status and Production At this point, we're now the second most frequently launched US rocket annually. We're increasing our launch cadence. Our launch cadence is really governed by our customers' readiness. As I mentioned before, we are a dedicated service, so we fly when the customer is ready. But we're certainly happy to see the launch cadence go up. The way I think about our launch manifest is it feels like a game of whack-a-mole. We have two rockets at the launch site at any time. It's very common for us to have two customers integrating. But as customers are continuously moving the manifest around to ensure we can achieve some level of cadence—if we didn't do that, then the launch cadence would be even less. So really, the number one driver for us at this point is customer readiness. Once you've got one away and a couple away, then the next big thing is scale-up. We have multiple factories—we have factories in five states in the US and we have three different shops down here in New Zealand. Scaling up is really the trick. I think our record right now is we built—the production line's record—as one rocket in 18 days. We're sort of there or thereabouts at the moment. Certainly every month there's a new vehicle that goes down to the launch site. The magnitude of what it takes to push a rocket off the production line every 18 or 20 days is really a huge collective and a finely tuned machine. One component that's not ready in time can hold up a test, or one test that doesn't go well holds up the production line. To get a regular production cadence, given the amount of testing that we do at every component level, subsystem level, and then full assembly level, is really an art. I think if you got a piece of paper and you wrote down all of the processes and everything that had to happen to get a launch vehicle coming off the production line, most people would largely arrive at the agreement that that's pretty improbable. So it's an incredible team that works incredibly hard to keep all of these launch vehicles rolling off the line. That's definitely the hardest part of the whole gig. ### Learning from Failure Just when you think you've got it all sorted and everything's going well, you get a baseball bat to the face. We got to flight 13 before we had our first failure. I was always told that flight one would fail, flight three would fail, and flight 10 would fail—that's the general rule of thumb. We got to flight 13, and the tiniest, tiniest, tiniest thing—kind of like flight one—it really does take the smallest thing in this industry to have a really bad day. I always tell customers, especially government customers, that there is nobody on this planet that wants to fail less than us. Mission assurance is a huge deal here at Rocket Lab. It's really hard to explain how devastating a launch failure is—not just for the customer, of course, not just for the company, but for the team. After flight 13, everybody was walking around here like it was a morgue. But we certainly didn't waste any time going to fix it. On flight 13, there's a high voltage connection that wasn't quite perfect. It passed all of the load testing and everything, but as the rocket was flying, the connection heated up. The most ridiculously unlikely set of events: the connection heated up, the potting compound around the joint melted and liquefied, flowed in around the actual joint to then isolate the joint, then the joint cooled, and as the copper bar cooled, it pulled across and re-contacted and sent a giant electrical spike—and that was the end of that. If you think of all the improbabilities of those things happening, it really is a very difficult thing—like I say, a baseball bat to the face. ### Reusability Program Onto something a little bit more cheery. I had to eat my hat for a couple of reasons. One is I said staunchly that a small launch vehicle would not be able to be made reusable and that we wouldn't build a bigger rocket. So the lesson there is never say never. For a long time, I really didn't see how a small rocket could be made reusable in the traditional sense, if there is a traditional sense, because there's just no margin. A small launch vehicle—the margins are just incredibly fine. To carry enough propellant to do a propulsive retro burn, let alone landing, is just absolutely not feasible. So we had to find a way to do reusability that let the atmosphere do the majority of the work. We don't have any mass to give this thing, so how do we do it? Electron's reusability process is basically pretty simple in concept but actually quite difficult to execute. As we separate, we reorientate the stage on its ballistic arc. We reorientate it engines first and we hold a very tight corridor as we re-enter. The blunt body of the first stage pushes the shock wave forward and enables us to kind of sit in behind the wake of all of the heat and the mess. There are some shock interactions on the very base of the rocket that are difficult to manage, but really the hardest thing is getting through that re-entry corridor and maintaining control. If you have a little bit of off-axis or change of angle of attack, basically the plasma knives come out and unzip the stage super quickly. Being able to control that and control that thermal environment with a composite rocket is very difficult. Once we're through that heating regime, we're able to go into a more traditional recovery where we have a parachute. Then, as some of you may have seen, the idea is to scoop the parachute with a helicopter before it hits the ground. I can remember when we first discussed reusability—I actually held an all-hands at Rocket Lab to announce the reusability program. I was full of excitement and bluster about how we were going to do it, and in my head at least, it sounded completely reasonable. I presented it to the entire company, and there was some giggling and snickering going on in the cheap seats for sure because it sounds a little bit crazy. But actually, it's relatively simple. The challenging bit is actually the rendezvous with the helicopter, not so much the collecting of the stage, but getting the stage to rendezvous. [Referencing video] It's quite cool—you launch 400 kilometers away on a ballistic arc, travel through Mach 7, Mach 8, and then in a period of space and time, rendezvous with a stage and snag it with a hook. It's pretty cool. ### Expanding to Satellite Manufacturing Launch was only the beginning. If you look at the very second kick stage that we ever flew, it had recesses in it for solar panels. We were always going to move into building our own satellites. We announced our satellite program in 2019-2020 and haven't looked back since. Just like scaling a rocket, we needed to scale satellite manufacturing as well because it's all about the scale of manufacturing. We're super lucky that we'd worked with some amazing people over the years and were able to bring them into the Rocket Lab family. The first being Doug Sinclair—you'll be pleased to see that Doug still has his couch over in the exhibit hall. Super lucky to be able to bring Doug's family in with our family. Then the team at ASI for software, and Walt and Mike at PSC—we've flown so many of their separation systems, best separation system in the world. And then a big pain point for spacecraft is obviously solar. Where all this really came together was when we did the CAPSTONE mission for NASA to the moon. This is where we not only needed a rocket, we had to build a spacecraft as well. ### The CAPSTONE Mission The trajectory here wasn't just a normal direct descent into lunar orbit. We had to build a rocket, of course, but then the Photon Lunar spacecraft spent eight days in orbit raising itself and doing this very unique trajectory maneuver to ultimately end in a TLI [Trans-Lunar Injection]. So a very complicated mission and spacecraft. We didn't have an engine, so we needed to build an engine to do this. This is a 310-second ISP, 100-pound thrust engine. For those who build engines, getting 310 seconds out of a tiny little engine is really difficult. Once again, this is an electrically pumped engine, but what probably is the best kind of application of electric pump—where you do the burn, you let the sun charge up your batteries, and you do another burn. It's really incredibly elegant. You need to carry almost no gas on board. But this engine was a huge development program. It's a non-toxic, green propellant engine at really high ISPs and reliabilities. Then we also had to build a spacecraft. The Lunar Photon is not a kick-stage variant of the Photon—it's a completely standalone thing. It's a three-axis stabilized spacecraft. It's a great big ball of delta-V, and actually in its own right, fairly complex—a sun-pointing spacecraft. There are just no margins here. The reason why there were no cameras on that launch is we just did not have the mass for cameras. Every gram was accounted for. The launch itself—Electron was originally designed to lift 150 kilograms to its sun-synchronous orbit. We're able to improve that and get that to 200 kilograms. This flight lifted 320 kilograms to low Earth orbit. So it's fair to say that Electron can lift 300 kilograms no problems at all, but that extra 20 kilograms? Really tough. This was the highest-performing Electron ever. The engines were at 110% the whole time, and we pushed that vehicle as hard as we could. The first maneuver was do-or-die. We separate off the lunar spacecraft, and generally if you've got a satellite, it's quite nice if you can have a little bit of time to commission it—maybe a couple of hours or even a couple of days to deploy the solar panels and stabilize it and slowly bring up systems. That wasn't the case for Lunar Photon. We had to separate that spacecraft off and liven it up and go straight into the first burn within minutes. So everything just had to work—all the sun pointing, the three-axis stabilization, the propulsion system, the TVC [Thrust Vector Control], the cold gas, star trackers, reaction wheels—everything had to work instantly. There was no gently bringing it up and commissioning the spacecraft. It was pop it off, and we're off. So very much do-or-die. As we were getting through the mission, we had originally a total of eight burns and TLI to do. The engine was actually performing—over-performing—really well. Every burn is risky, and every burn we're doing in the dark, so we decided to do the super maneuver where we combined burns six and seven together to give us a total of a kilometer per second of delta-V in one burn. You can see in that diagram, the first white line was intended to be the first burn, and then the other white line was the second burn. We did it all in one hit. There's nothing like a successful mission. We did the TLI, and it was right down the middle of all the requirements. The TLI is not just a position in space—it's a position in time to enable the rendezvous. This was a seven-day mission, 24/7 operations on the spacecraft, and incredibly accurate orbital insertions and TLI insertions are required. So huge success for the team. Lunar Photon now is kind of stooging out there in our solar system. It's around about 1.3 million kilometers away from Earth. It's on a free return trajectory past Earth later this month. As it scoots past Earth, we've still got about 10-15% residual propellant in there. We'll have a crack at doing something cool with it and see how far we can get into our solar system. So the spacecraft as of today is happy and healthy. It was only ever designed for eight days, but we're using this opportunity to learn what it's going to take to get to Venus and other far-off destinations. ### Looking Forward How do you try and put 16 years of work in a 30-minute talk? You can't. There's a whole bunch of exciting stuff going on. We're going to be giving a Neutron update in the middle of September, so keep an eye out for that. That program is going really well, and there's lots of exciting stuff to talk about there. In our Space Systems division, the next off the block is two missions for NASA to Mars—the ESCAPADE spacecraft. We have commercial missions for Varda, where we're responsible for the spacecraft and re-entry targeting of their on-orbit factories. We've got the Venus mission and heaps more exciting stuff going on.