New Space Revolution – Size Matters

This is the fourth in a series of articles about the current state of the space industry. If you have not read the preceding articles, you might find it useful to do that before reading this one as it builds on the arguments made in the earlier parts. This week, company size matters. I ended the last article by concluding that partly because of the response of "traditional" space companies and of governments the space sector is now divided more on the basis of size than on the basis of whether the companies involved are "new space" companies or "old space" companies. Many large traditional space companies are involved extensively in producing spacecraft that a few years ago would have been seen as "new space" vehicles. They are smaller, they are much cheaper, and there are many more of them. The customers for these satellites are no longer governments. However, the projects themselves are very large - often comprising hundreds - even several hundreds of satellites. And while the customers are private sector entities whose business models revolve around sales to a wider market - governments are significant stakeholders in many of these projects through a variety of support and subsidy programs. In effect, this has also meant that some significant "new space" players have also grown and migrated to this end of the market. And while they continue to operate to business models that are different from the traditional space industry - drawing their funding from private sector investment - and aiming their offerings at wider commercial markets rather than a countable number of large reputational risk averse customers - they have become much larger players and many have developed close relationships with the traditional space primes. This matters. It matters because with this scale and with these relationships comes access to expertise. Expertise and experience. Access to expertise and experience that is much harder to come by for smaller space companies. Which is why size now matters in the space business. It is this fact which is driving the fragmentation of the industry at the lower end of the scale. It is why this fragmentation generates both threats and opportunities that I would argue will eventually drive the consolidation when it comes. But let's back up for a little bit and examine the relationship between scale and access to talent - and why this relationship is particularly important in the space business. In any business sector there is always a relationship between the ability to attract talent and business size. This is always one of the dynamics that plays out when entrepreneurs arrive seeking to disrupt existing businesses. In fact, often the advantage of the entrepreneurial players is that they offer a more attractive environment for certain kinds of talent. Startup companies have a reputation for being less structured, for providing more opportunity for innovation and more opportunity for more exploration of the "risk - reward curve" by offering employees a less obviously secure future in return for meaningful participation in the fruits of success if and when it comes. Such a culture, in turn, tends to be less attractive to more experienced employees. Which is not to say that there is no movement between established companies and startups. There is. In fact, founders of startups are often former employees of large established companies. But, it is also certainly true that more experienced employees are less likely to embrace entrepreneurial culture for a lot of reasons - from decreased tolerance for risk that comes with increased responsibilities at home to comfort with their existing corporate culture and systems and a clear career development path within that culture. So, not unnaturally, the employee base at small entrepreneurial companies is often much younger than it is at large companies. It also means that this employee base comes with less experience. This certainly seems to be true of the space sector at this time. Now, this lack of experience can be seen as a source of strength in that it fosters a greater degree of innovation and less aversion to taking potentially disruptive risks. In some ways it is this very demographic fact which promotes the "fail early, fail often" culture which is seen as the critical philosophy that gives entrepreneurs their advantage over established players. But, in the space business there is a bit of a catch that may not be true in many other sectors. And that difference arises because it is: the space business. Space, in and of itself, provides an environment that is both extremely demanding and hard to simulate. Which means that the experience of having worked on actual spacecraft is irreplaceable. Such experience is also hard to come by. To explain further, it's worth looking at some particular kinds of expertise that working in and from space requires in some detail. It's not my intention, here, to make a fully comprehensive list of skills needed to go to space, but I do want to give some examples of why depth of skill matters more than breadth of knowledge in some important ways. For instance, let's consider thermal and mechanical engineering for space. At first you might think that this is not that complicated because space is just cold - and so you just need to have a design that can cope with the cold. But this is not actually true. Space itself is not actually cold. Because there is no atmosphere in space, the environment itself is actually neither warm nor cold. In fact, the heating or cooling experienced by an object in space is very much determined by its exposure to radiative sources of heat. Of course, the principle one of these is the sun. Point a thermally absorptive surface in space at the sun and it can become very warm indeed. But, point a thermally emissive surface away from the sun and it will rapidly remove all of the heat from the object it is attached to. So, for instance, a small satellite orbiting the Earth in a low Earth orbit may experience extreme cycles of heating and cooling every 90 minutes or so as it passes in and out of the daylight side of the Earth. On the other hand a satellite in sun-synchronous orbit will experience constant heating. While a large satellite in geo-synchronous orbit may experience both extreme heating for surfaces pointed at the sun and cooling of those surfaces pointed away from it. The impact of all this heating and cooling is quite different in a vacuum than it is on earth. In a vacuum heat must be conducted away from hot spots and toward cold outer surfaces where it can be radiated away. But if those same surfaces are hot, the heat will instead be conducted into the spacecraft instead. This means, in general, that many parts of the spacecraft are likely to experience ranges of temperature and rates of temperature cycling that are unlike anything that will be experienced on Earth. Of course, in space you also have to add to that the consideration that these cycles will be experienced in an environment without gravity. For mechanical devices this can be very significant. It is partly why spacecraft engineers still have, after 60 years, a reasonably strong allergy to relying on mechanical "mechanisms" to accomplish critical functions. This is particularly true for mechanisms that have to operate repeatedly or to very fine tolerances. Such mechanisms can be, and have been designed - but it is always an experience fraught with a mild dose of terror for the designer. And this is because it is virtually impossible to physically simulate the environment of space without going there. Of course it is possible to simulate thermal cycling. It is also possible to place objects in a vacuum. With enough expense it is possible to do both at the same time. Similarly it is possible to subject objects to periods of micro-gravity - but only for small periods of time. Trying to do all three at the same time? I suppose that could be done. But it's not. Mostly because it would be entirely cost-prohibitive. Finally, consider the fact that by the time any of those mechanical parts reach that hard to simulate space environment they will have had to survive a space launch - which still consists of strapping a sensitive instrument to a bomb with a hole in one end and giving it the ride of its life for 5 minutes or more. Again, all of that can be simulated. And again, all of that physics is well understood and can be taught. But the process is complex enough that there really is no replacement for having done it. More than once. Preferably much more. As another example consider the radio frequency engineering of a spacecraft. First of all, bear in mind that RF engineering, as it is called, is still considered by many - especially other electrical engineers - to be a black art. A subject that is really only understood by those who, by dint of long practice have been made privy to its arcane mysteries. Then consider that the very nature of spacecraft is that it is a long way from home - always - and that it is really only useful if it can speak to and be spoken to from the ground. This means that it must contain - at one time - a fairly powerful RF transmitter as well as a fairly sensitive RF receiver. And, of course both of these functions must be hosted in a very limited volume. That alone requires very careful design to ensure that the sensitive receiver is not continually swamped - or even damaged - by the powerful transmitter. But, wait there's more. Consider the fact that many spacecraft contain very precise instruments. These instruments invariably depend on electrical components. Those components are likely to be susceptible in some way to RF emissions (such as those produced by a transmitter talking to the ground). Those components may also very well emit RF radiation - which even in small quantities may have the capacity to interfere with the aforementioned sensitive receiver that is listening to the ground - or the other parts of the sensitive instruments that are hosted on the spacecraft. Again, all of these effects are well understood. The techniques for calculating them can be taught. Test facilities to test the effects exist and designs can certainly be tested. But, as I said, RF engineering, at the best of times, is a discipline that involves a lot of subtlety. The ways in which RF signals travel through various materials is predictable, but when those signals encounter junctions and changes in geometry the results can be surprising - especially for someone who is not experienced in that analysis. The people experienced in that analysis are still relatively few in number. Those that have done it in the context of having to also design a system that is capable of surviving the aforementioned thermal and mechanical environment are pretty rare indeed. And these are just two of the engineering disciplines involved in building the physical spacecraft. There is another, perhaps more insidious feature of working in space that makes experience a valuable - and still rare - quantity. And that is the fact that when you run a business from space, the infrastructure of your business is completely remote from you - and always will be. Once your spacecraft or sensor has been launched there is very limited opportunity to update, upgrade or even maintain it. And what capacity there is to do any of these things has to be planned in advance. Which means that everything that you do from space - which is the basis of your business - has to be planned pretty much years in advance. In very real terms, ultimate success in a space-based business requires the ability to plan a long way into the future. Such plans need to include not only basic business strategy, but may also need to include some fairly detailed engineering design of what will eventually become the beating heart of your business as much as 5 years in the future. And this is where the entrepreneurial ethos of "fail early, fail often" collides with the reality of running a business in and from space. The statement about failure is meant to convey the sense that risk aversion is often a barrier to progress, that it is more efficient sometimes to build and test new ideas - even if they are unsuccessful - than to spend too much time examining how they might fail. It is an entirely valid approach to solving problems. But, it is predicated on the assumption that every failure provides an opportunity to learn and that every design iteration builds on the previous one. The problem in working in space is that there are many ways to fail that are not instructive at all. For instance, say a startup company builds a new and innovative sensor for examining the Earth. A custom designed satellite is built around it and it launches successfully. Getting this far is certainly a success, but there are many ways in which this mission could still fail - but which would not do very much to advance the technology or the business. Something as mundane as a failed ground station which does not contact the satellite when it is deployed from the launcher could mean that the company never talks to the satellite because once it leaves the launcher its location becomes less and less predictable. The satellite could fail to separate properly from the launcher because of a poorly designed deployment mechanism, or commercial grade parts in some unrelated piece of electronics might off-gas under vacuum and coat sensitive optics, or the detector electronics may turn out to be susceptible to the RF radiation of the communications antenna. All of these failures may seem to embody the philosophy of "fail early, fail often" philosophy of just getting a prototype in service so that it can teach you how to design the next iteration. But some of them will not be instructive at all because the satellite may never actually operate. Worse, some of these kinds of failures may just repeat mistakes that have been made by others in the past and for which established solutions or mitigation procedures are available - meaning that while such a failure represents a learning opportunity, it's not a particularly cost-effective one compared to having access to existing knowledge and expertise. Because, to put it bluntly, launching a satellite is still an expensive way to learn how to design one. Even with the lowered costs of access to space. Access to design expertise and experience is still often less expensive than launching a satellite. Much less so. It is not my intention here to claim that the only successful way to design space hardware is the tried and true method of the traditional industry. It's not. Every day companies are proving that there are smaller, faster, and cheaper ways of getting to space. But also, every day, companies are discovering that getting to space is still harder than it looks, and that it is possible not to fail early and often, but merely to fail entirely for want of a critical piece of expertise or experience. Which brings us back to the point of this article. The expertise and experience needed to get to orbit and operate there reliably and effectively requires specific applications of a variety of very technical disciplines. In depth expertise and experience in these disciplines is only really available by going to space. This means that there is a strong economy of scale in the business of designing and operating spacecraft. It is far more efficient to hire a single expert and spread their experience across multiple projects than it is to hire all of the expertise that is needed and apply it to a single satellite. Which is why simply reducing the cost of launch has not actually reduced the cost of access to space by as much as you might think. It is, also, to some extent, why until 10 years ago the main players in the industry consisted of large "horizontally integrated" companies that could afford to develop and maintain the niche engineering expertise they needed because they could spread it across multiple projects and make it cost-effective. In short, it's why, in space, size really does matter. And it's why the bifurcation of the market in large players, with large projects, and small players, with large ambitions matters as well. In the next installment of this series we'll talk, in particular, about how the fragmentation at the small end of the space business represents both an opportunity and a threat.