Op-ed | NewSpace 2.0: Moving beyond the Minimum Viable Product
This op-ed originally appeared in the May 14, 2018 issue of SpaceNews magazine.
NewSpace is a term coined to describe the visionary startup companies that are using satellites to address audacious questions like: What if we could take detailed images of any spot on the Earth, once every hour? What if we could use satellites to connect the ‘other’ three billion people who don’t have internet access? What if we could return valuable minerals by mining asteroids? In the so-called Old Space paradigm, questions of this magnitude would be addressed by national governments in multi-decade studies which — let’s face it — would probably end up de-funded before reaching conclusion. Things have changed. Owing partly to the emergence of small satellites, NewSpace startups are disrupting commercial space by addressing these and other questions on shoestring budgets compared to government programs.
The thesis of smallsat philosophy is that many small satellites can be launched much, much cheaper than traditional schoolbus-sized satellites and, despite their small size, can collectively deliver data with quality comparable to their larger and more expensive predecessors. NewSpace startups are attempting to prove this thesis using Eric Ries’ principle of the Minimum Viable Product: “The first step is figuring out the problem that needs to be solved, and then developing a Minimum Viable Product (MVP) to begin the process of learning as quickly as possible. Once the MVP is established, a startup can work on tuning the engine.” An MVP consciously omits features that are necessary for longterm business viability in favor of proving or disproving critical challenges early on.
A growing number of NewSpace companies can now claim successful MVP rollouts. Planet and Spire have deployed more than 400 small satellites that prove it is possible to send satellites up, send data down, and do this without breaking the bank. Iceye demonstrated the ability to use a small satellite to image the Earth through clouds. Telesat LEO is proving small satellites’ capability to provide low-Earthorbit communications. By all measures, these early NewSpace deployments are impressive and disruptive.
As awe-inspiring as these early NewSpace exploits are, it is even more impressive to realize that we are only witnessing the MVP version of these architectures. In the immortal words of Bachman Turner Overdrive, “You ain’t seen nothin’ yet.” The next leap in smallsat mission utility will not come from a revolutionary new sensor or payload, but from the careful and considerate placement of each vehicle into a precise orbit that empowers the entire constellation to have value greater than the sum of its parts.
Consider how smallsats are deployed now: a large number of satellites share a ride into space on a single rocket (the record is 104). Upon reaching orbit, the rocket deploys the ‘flock’ of satellites, which then drift in a rather uncontrolled fashion about the Earth. For an imaging architecture the flock approach is functional, but not optimal. Sure, each satellite can take pictures and send them down, but individual vehicles may be bunched together and thus return redundant images. For a communications architecture, a flock is not functional; random gaps in coverage that naturally occur in an uncontrolled group would be unacceptable to the end user. But even with these limitations, deploying an MVP is an essential step in the build-measure-learn loop.
The value of smallsat architectures will be fully realized when the disorganized flock is instead transformed into a carefully orchestrated constellation, where evenly spaced smallsats are spread out to optimize their coverage and data value. The need to evenly distribute satellites in a constellation has not been overlooked by the NewSpace pioneers. Instead, this feature was consciously omitted from MVP rollouts in order to focus first on bigger challenges. With those challenges on the run, constellation optimization is now in full swing for next-generation vehicles.
There are two ways to implement constellation optimization. The first is to launch smallsats individually, or maybe two to three at a time, on specialty small rockets that can deliver spacecraft “where you want, when you want.” Because the smallsat passenger is paying the full ride fare, it is free to pick the destination. A sequence of these boutique launches can then populate a constellation of arbitrary configuration. Companies like Rocket Lab and Vector Space Systems are leading this charge by chauffeuring smallsats to private destinations. There are two challenges to this approach that may not make it suitable for all constellations. First, large constellations would require a large number of launches — even if you launch one rocket a week, it could take months or even years to fully deploy a constellation. Secondly, on a per-kilogram basis, the boutique launches will be more expensive than the large rockets that are used for rideshare, so the total deployment would be higher.
A second way to optimize a constellation is to equip each smallsat with on-board propulsion. Many vehicles can then share a ride to orbit and enjoy the low per-kilogram costs of launchers such as SpaceX’s Falcon 9 or the Falcon Heavy. Although all of the vehicles will be dropped off in a big bunch, they can use their individual propulsion systems to disperse into individual orbits pre-chosen to optimize constellation uniformity. This approach comes with the added benefit that the on-board propulsion system can be leveraged to provide additional mission value — for instance, to extend mission lifetime by compensating for drag, to re-configure a constellation in response to a node failure, or to de-orbit vehicles at the end-of-life to make room for replacements.
Until now, NewSpace companies have not fully leveraged on-board propulsion for their smallsats, largely because the technology hasn’t matured enough and the costs have been too high for inclusion on an MVP. The types of propulsion systems necessary for this deployment model have historically carried multimillion-dollar price tags, an expense level that isn’t tolerable for or conducive to the NewSpace model. To complicate matters even further, the types of propulsion systems necessary to make this approach viable and cost-effective have historically been built one at a time by hand. The design, development, and manufacturing processes have not evolved to produce the necessary assets at scale. In order to move smallsats beyond MVP there is a market need for on-board propulsion systems that can be manufactured in large quantities, at prices far below currently available offerings.
More than 3,600 smallsats will be launched in the next decade and you can bet that they won’t all be in one big flock. Over the next few years, expect to see each and every smallsat launched either via chauffeured custom ride or with its own on-board propulsion system. With the era of NewSpace MVP moving into our rear-view mirror, it will be exciting to see what is down the road. New(er) Space? NewSpace 2.0? Whatever you call it, it will be exciting.
Brad King is CEO and co-founder of Orbion Space Technology, a Michigan-based supplier of electric propulsion systems for small satellites.