The SmallSat Launcher War

Over the last decade or so the definition of what a ‘small satellite’ is has ballooned beyond the original cubesat design specification to satellites of 50 or 100 kg. Today a ‘smallsat’ is defined far more around the cost, and sometimes the technologies used, than the size and shape of the box that goes into orbit.

There are now more than fifty companies working on launch vehicles dedicated to lifting these small satellites into orbit, and while nobody really expects all of those to survive the next few years, it’s going to be an interesting time in the launcher market. Because I have a sneaking suspicion that Jeff Bezos’ statement that “there’s not that much interesting about cubesats” may well turn out to be the twenty first century’s “nobody needs more than 640kb,” and it’s possible that everybody is wrong about how many of the launcher companies will survive in the long term.

Building Smaller Satellites

Small satellite builders are using a lot more off the shelf technology than the old school aerospace companies. While that inevitably means more on-orbit failures, the cost of production of a small satellite is so much lower than a traditionally sized payload that it’s actually an acceptable trade off. Off the shelf technology also means that the barriers to entry to building a satellite are considerably lower and development times that much shorter.

Planet Labs Dove 2 Satellite
Planet Labs Dove 2 Satellite

In an industry that typically takes decades to build and launch a satellite the ‘new space’ companies like Planet Labs who, starting in a garage, have built a constellation of over 150 Earth observation satellites in five years, comes as a serious disruption.

When 26 of the Planet Labs “Dove” smallsats were destroyed  in October 2014, with the loss of the Antares Orb-3 mission, the company built several replacement satellites, and had them tested and delivered to NASA in just nine days. Those satellites were successfully launched in January 2015 on board a SpaceX Falcon 9. That sort of turnaround time is industry breaking, and the model of how launches are bought and sold will have to change to accommodate payloads built on much shorter timelines. The typical two to three year timeline from contract to launch, used by most traditional operators launching to GEO will not work for most of the new space companies.

Making Launchers More Available

Launch of a PSLV by the Indian Space Research Organization
Launch of a PSLV by the Indian Space Research Organization. The PSLV is scheduled to launch 104 satellites on 15th Feb, 88 of them will belong to Planet Labs.

The recent availability of low cost piggy-back opportunities on board medium, and heavy-lift launch vehicles, has attracted small satellite payloads in rapidly growing numbers. But right now most, if not all, small satellites are launched as secondary payloads to larger, and far more expensive, satellites.

There are still unused capacity on board those heavy launchers, so the problem is absolutely not launch availability, at least not yet. The problem for most small satellite builders is whether the launches happen when they expect, and the limited choice as to final orbit for their payloads.

However while launch availability isn’t a problem right now, it may well become a problem soon as the availability of Falcon 9 secondary payload opportunities dry up. Just over two years ago now Elon Musk walked on stage and announced that SpaceX was opening a Seattle office dedicated to designing and building the 4,000 small satellites that will make up a Low Earth Orbit constellation that will provide an Internet connection anywhere on the planet, and as a byproduct, provide funding to support the SpaceX core mission, to go to Mars.

Building Constellations of Satellites

You might argue that satellite broadband already exists, and isn’t particularly profitable. You’re right. It does, and isn’t. But the satellite constellation Musk is building is somewhat different. Current satellite networks are based around geostationary satellites in high Earth orbit, with correspondingly long communication delays, and a round trip latency of around 600 ms. In contrast Musk’s satellite constellation will be in much lower orbits, with altitudes ranging from 715 miles to 823 miles, and much smaller latencies of between 25 and 35ms, similar to existing ground-based networks.

Keynote at the 2017 SmallSat Symposium shows a projection of constellations put into orbit. This photo was tweeted by Erik Franks
Keynote at the 2017 SmallSat Symposium shows a projection of constellations put into orbit. This photo was tweeted by Erik Franks

Musk projects that, while ‘last mile’ connection for most people will still be fibre, the SpaceX satellites will be able to handle 50% of the global backhaul capacity, and they actually expect to pick up most of that traffic.

You might call that crazy except that Musk, and SpaceX, isn’t alone. They weren’t even first. OneWeb has similarly proposed a constellation of around 700 satellites, and while they have a slightly different approach they’re also chasing the profitable backhaul market. The 21 Soyuz launches they have purchased to put their constellation into orbit has been called “the biggest commercial rocket buy in history.”

If they, or SpaceX, are successful your Internet connection might well be carried over their satellites without you even knowing about it.

Are We Going to Space Today?

While the last few years have been filled with announcements, and testing, there have been few launches. But that’s about to change. At the tail end of last year Rocket Lab announced the completion of their launch complex built on New Zealand’s Mahia Peninsula. The company’s two-stage Electron rocket is designed to carry payloads of up to 330 pounds (150 kilograms), and they’re on track for test flights early this year, with commercial launches starting in the second quarter. A single dedicated launch of the Electron is priced around $5.5 million, with the price to orbit for a 1U cubesat around $50,000 to a 500km sun synchronous orbit.

With four, possibly five, companies looking to make their first launches this year, including Rocket Labs and Virgin Galactic’s Launcher One, the competition to be first is heating up. At least at the moment, Rocket Labs looks to be winning the race to be the first commercial smallsat launcher. Although they still face heavy competition, and not just on scheduling. With commercial smallsat launch companies now entering the market from China, pricing too is also coming under pressure, with the price to orbit for a 1U cubesat potentially dropping to around $10,000.

Next Stop, the Moon?

[tweet 829091801344544769 width=’400′ align=’left’]

Of the five Google Lunar X Prize finalists with secured launch contracts, two are riding as primary payloads onboard the new generation of small satellite launchers. With Moon Express, the only US based team to make it to the final, having signed the world’s first multi-mission lunar launch contract with Rocket Lab for 3 lunar missions. The first is scheduled to launch later this year, numbered amongst the first few launches of the new Electron rocket.

Whatever happens the next year, or possibly two, will be make or break for both the new launcher companies and the smallsat builders that want to ride with them. If they succeed the price of launching your own space program is going to drop dramatically, and bring the cost of having your own satellite into line with buying a mid-range car. At which point there’s probably a lot of people reading this that might well decide that having a satellite is cooler than having a car.

54 thoughts on “The SmallSat Launcher War

  1. “Because I have a sneaking suspicion that Jeff Bezos’ statement that “there’s not that much interesting about cubesats” may well turn out to be the twenty first century’s “nobody needs more than 640kb,”

    Almost guaranteed.

    1. It’s not really comparable.

      The hoopla about cubesats is more like, “These 640 kb memory modules are going to be the bees knees in the future, mark my words!”

      Of course microsatellites are going to be a thing, but there’s not much you can do with such a tiny object because of power constraints.

        1. Sure but anything interesting in a cube sat likely needs to talk to the ground and the energy requirements for that aren’t that likely to improve anything like processing power or processor efficiency.

          1. I believe the next gen of Iridium supports communication from sats in LEO (Ir satellites are well above LEO). It might not be fast, but small and relatively cheap for applications not needing high bandwidth.
            I expect that if small sats do take off, we could see more interest in the satellite to satellite market.

          2. I’m sure we’ll see small satellites built as “modules” and assembled in-orbit. Launch your power and telemetry sat first, and rendezvous with an experiment module, propulsion module, whatever else, later. Not as efficient as a one-launch satellite but if/when the industry takes off we’ll probably see more of this kind of modularity.

          3. Given that Bezos and others intend to build what is essentially the internet of space up there, future cub sats won’t be communicating directly to ground.
            They will transmit to the space bound internet, to ground via the mid sized satellites.
            Cubesats BECOME useful precisely because they won’t need to worry about the com channel that much. Synergy in action.

        2. The power of a cubesat is constrained by physics, not by technology.

          Consider, there is only so much sunlight hitting a 10x10x10 cm cube. Conversely, if you put some sort of nuclear battery inside it, there’s only so much heat it can radiate out to keep itself from melting. You simply run out of space and energy to do a lot of interesting stuff.

          1. “The power of a cubesat is constrained by physics, not by technology.” that’s a very risky assumption and i can prove you wrong in a second (unless you specifically mean current tech then your right). even if i work with your assumptions both nuclear and photovoltaic technologies are rather inefficient at the moment. imagine you could through a leap in technology now make solar cell or nuclear batteries that are 95% efficient that would boost available power by a huge margin. not to get all nerdy on you and talk about ZPM’s and Naquadah generator but what if we finally crack fusion power 10-20 years down the line we may be able to create micro fusion reactor that would fit nicely in a cubesat… who knows….

          2. @Dax – Without accusing you of being myopic, consider that the applications wanted themselves will have changed as well – Smart Dust type technologies may not require that each node do all that much and thus would not need a great deal of energy. Of course I have no idea if this will be the case, I only caution against making sweeping statements about power requirements being a hard constraint to development because history shows that this rarely the case over the long haul.

          3. >” imagine you could through a leap in technology now make solar cell or nuclear batteries that are 95% efficient”

            Yes, I can also imagine if cows could fly. There’s also physical limits to solar panels and the efficiency of RTGs. Carnot’s law comes to mind – which is why I brought out the heat dissapation problem.

          4. >”Smart Dust type technologies”

            Is mostly scifi. It’s at the same level of imagination as intelligent nanobots coursing through your veins. When you’re the size of a bacterium, you can’t be much more complex than a bacterium.

            >”Maybe the ittybittySat {TM] will have a wire loop trailing behind and harvest power from the Earth’s magnetic field.”

            They would do so at the cost of their own momentum, which would make them fall out of orbit.

          5. “Is mostly scifi. It’s at the same level of imagination as intelligent nanobots coursing through your veins.”

            Your point? The issue here is not making myopic pronouncements about what the future may bring, because that has been a foolish position to take. A computer in my pocket was SF within MY living memory.

            When you’re the size of a bacterium, you can’t be much more complex than a bacterium.

            Well those organisms can do quite a lot, especially in number. I wouldn’t be so quick about writing off the potential impact of such things.

          6. It appears you don’t work with satellites. We use deployable panels with solar cells. We also don’t use 1U form factors for commercial satellites. It’s more like 6 to 16 Us. Still not huge amounts of power, but plenty for radio from LEO. The industry is moving *fast*!

          7. @Logan

            There are already solar cells that exceed that efficiency limit. (The Shockley–Queisser is 33.6% but applies only to single-junction cells. The current record according to Google is >36%.) This uses a multi-junction cell where the blue light is is eaten by the first layer of the cell, but the red light passes though to the second layer (or vice-versa) so you don’t waste the extra energy of blue photons to get something that can harvest red photons.

            And of course a (10 cm)^3 cube at launch doesn’t have to stay in that configuration, depending on hw extreme you want ot get with deployable solar panels.

          8. @Ren sorry clicked Report when I meant to reply. Using energy harvesting like that is actually converting the kinetic energy of the sat into electricity and will eventually cause the satellite to de-orbit.

        3. Part of the power constraint problem can be dealt with by spending more on the ground equipment.
          Using a 20 foot foot dish to talk to it wouldn’t add much to the total cost of the mission.
          The first satellites such as Explorer 1 and Vanguard sats would be considered nano satellites today.

      1. Did you see this in the article?
        If planet lab can build out a network of cubesats to image the entire planet daily, there’s a lot more the technology can do than you think.

        Here’s an article talking about people who are trying to use cubesats for further out missions:
        One possibility might be for a cubesat that can inspect Saturn’s icy rings. A larger craft can’t approach it without risking damage. Even mentions chipsats that can attach to individual ice particles.

        1. >”A larger craft can’t approach it without risking damage. ”

          I don’t see why getting hit by speeding debris would hurt the small satellite any less. The problem is that the spacecraft is traveling at such a large speed difference to some of the stuff flying out there that any collision would be like getting shot with a howizer.

          Even as you accelerate to the average orbital speed inside the rings, the stuff in there is flying about and colliding with itself and generally sandblasting your satellite away.

          1. The main advantage of small sats in a mission like that is they are cheap and light so you can send a lot of them and and be able afford to loose a few.
            If you sent 50 small probes it doesn’t matter if you loose 30 of them if 20 survive to send the data back.

  2. Remember that scene from Wall-E where a ship leaving earth pushes its way through a layer of satellites. Yeah that. On another note I am really excited to hear about the possibility of satellite broad band with < 50ms rtt. That will ensure internet for the world comparable to what GPS did for navigation around the world. Exciting times.

  3. So tiny satellites become as ubiquitous as small computing platforms that are in everything we touch?

    Problem is that all the junk in our phones/computers/thermostats/coffeemakers isn’t orbiting the planet at 28,000 kilometers per hour at approximately the same low earth orbit altitude (which will be uniform because of launch cost per kg.). Tiny satellites will be, and despite the general assertion, Space at a particular elevation above the earth isn’t infinite.

    The Earth may be the first planet we know of that has an artificial shell (or rings, if we can scare up some very dense artificial satellites to act as shepherd moons) of consumer-fad electronics. Given the rare-earth materials in most of them, the meteor showers should be very colorful, though.

      1. You think only in the present…

        What if these tiny satellites can harvest energy not only from the sun, and significant temperature differences on their surfaces, but in addition from the Earth’s magnetic field using conductive tethers? Then these “nanosats” may become permanent LEO’s. This raises more questions: (1) If the Nanosats are persistent who will manage them? (2) If there are swarms of these Nanosats in LEO (even if they are functional) don’t they amount to pollution? (3) When these self-powered myriad Nanosats occupy similar LEO orbits, what/who will prevent them from colliding and creating debris to pollute the orbitals? This is an important question for small satellites that use tethers that are many kilometers in length. When two Nanosats with long tethers begin to approach each other, the electromagnetic fields between the two interact making orbital prediction/steering too complex to manage.

        On the flip-side, if Global Warming (man-made, not man-made, or a mix of the two) seriously affects the Human population on Earth, it may be possible to use these self-sustaining Nanosats to form a “Blanket” around the earth reducing sunlight. A good solution? IMO no, but given how the science of Global Warming has been politicized to the point where real Science is no longer discernible, it may be our only option in the future.

        Another approach is to let these Nanosats proliferate unregulated, but have some Global Government body (UN – I don’t think so) just “Mop-Up” these things with near zero consideration to where the came from and what they’re for?

        No matter where this subject of Nanosats goes, you can be sure of ONE THING. The public discussion will be useless because of Political and Commercial self-interests who will pollute the Science with fakery. This is what is happening today on both sides of the Global Warming discussion. Expect more of the same – it is Human-Nature :-(

        1. Japan just attempted to test a tether system to gather energy from orbit. The test didn’t unfurl correctly, so it failed. Their idea was to use it to deorbit space junk. Because the process of generating electricity in this manner causes the velocity of the craft to slow down so that it re-enters the atmosphere. You can apparently feed energy into it to boost your orbit though, which may be what you were referring to.

          1. The tether was supposed to drag the craft down to de-orbit it.

            The idea of collecting energy from the magnetic field doesn’t work because the energy you collect comes from the orbital momentum of the craft itself. You’re turning its kinetic energy into electricity as it travels in the magnetic field, which causes the craft to lose orbit.

  4. And yet I can’t help but think: “Kessler syndrome”. Especially for higher orbits that don’t decay all that fast and small sats that have no active deorbit option. The on-orbit recovery/cleanup business is going to be booming too. Time to call in Roger Wilco.

  5. Is there a satellite network yet? I can picture backhaul for private entities, but I am thinking of something similar to fidonet. You could link to one sat, and find your destination sat via routing and addressing. Ideas are easy I guess.

  6. The cool part for me is that when I was a kid growing up in Napier, New Zealand, the Apollo program was happening and it was just awesome – I was a total space enthusiast. Today, just across the bay, NZ is preparing to launch international satellites with rockets developed largely locally (US support was needed for testing and commercial funding) – and one of the earliest is going to send something Moonwards. Who’d have ever imagined that our nation, known through the twentieth century for its pastoral production, would ever do this? There’s also the economic benefit for New Zealand which has been estimated at around potentially $500 million or more over the next while, which underscores the scale of the growing small-sat market worldwide.

  7. If I have done my math right (very possible that I haven’t), the red thing in the headline pic is supposedly an electric motor the size of a soda can generating ~100HP!! That kind of power density seems exceptionally high, but if there’s any explanation I can think of is that it has plenty of liquids to use for cooling.

  8. 1. Cube Sats are useless for commercial uses.
    a. Solar power collection is geometrically based.
    b. Bandwidth from LEO is a function of distance (orbit), radiated power, and cone angle (coverage).
    c. The power and speed efficiencies of making things small (e.g. ICs) has no bearing on transmission. i.e. if one kilowatt is required, one kilowatt is needed. Making it small doesn’t affect the power requirement.
    d. Small Sat’s orbits are poorly defined, largely uncontrolled, and degrade rapidly.

    2. Cube Sat does not equal Small Sat.
    a. Cube Sat is a form factor. ~10kg or less.
    b. Small Sats are roughly 150kg to 2000kg
    c. Small Sats have the ability to do commercial work due to their size threshold.

    @msat – Yes, we question the viability of Rocket Lab’s electromotive propellant pumps.

  9. With the dramatically lowering price for hauling stuff into space the engineering of the payloads is also changing.

    Here in Berlin there is a small company doing flight hardware and they say: why hire an engineer who works a year to reduce the weight without impacting structural strength when I can just launch the extra mass for half the price? Their tagline: the best against aluminum is more aluminum.

    This indicates an ongoing paradigm shift.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.