The 2018 National Defense Strategy gives out a set of goals and they all say “we want you to modernize in an agile manner with no excuses.” So that’s how we developed SDA [Space Development Agency]. We said, “what do we actually need to do to be able to launch satellites on two-year time frames, very similar to commercial software development.”
That was Derek Tournear, assistant director for space in USD(R&E), “Space Development Agency seeks input on satellite plans.” The SDA recently put out a request for information (RFI) on a new space architecture. As Tournear said, the SDA doesn’t have the legacy operations of the Air Force, and so can explore new concepts from scratch.
The RFI outlines an notional architecture of seven layers of satellites: space transport, tracking, custody, deterrence, navigation, battle mgmt, and support. The way Tournear speaks in the interview, it seems like those seven layers are pretty set in stone. Here is where they are looking for input:
The hardest part is actually the cross-links to make this interconnected mesh network in space. So we’re looking for technical solutions to enable that in an affordable manner.
So SDA is teamed with a lot of the services and other government agencies to shape their solicitations. So the next thing you’ll see will be coming out of the USD(R&E)’s office to have a solicitation for this comm backbone layer. That’s the first thing.
Before having seen the interview with Tournear, I asked MIT’s Bruce Cameron on the Acquisition Talk podcast for his advice on the SDA’s RFI for a new space architecture. He seems to have nailed it:
This is a very broad use of the term architecture. One of the alternative definitions of architecture is a collection of interfaces that imply something about interoperability. And I think that the way this is framed it looks a lot more like a collection of interfaces to me.
… one of the useful things to think about is what are the interfaces we are going to be strict about? In the position paper put out by the DOD there is an expression of a bunch of these layers, and that’s obviously a metaphor in the sense that these aren’t explicitly going to be layers, and like any good layered architecture I would very much doubt that these would have just a single interface between layers…
But I think it really does beg consideration of what we think the opportunity for the consistent enforcement of a couple of interfaces is, and what would those interfaces be. So I would spend some time on the interface problem, as opposed to shiny new hardware.
Still, whereas the layered architecture seems to be set by the SDA, Bruce recommends that many alternative architectures be explored, not just in terms of interfaces, but in terms of desired capabilities. He doesn’t advocate adding more and more decisions to a single architecture as though it were the default. “There is a lot of power in exploring the trade space before you go and down select.”
Leave a Reply