By COLIN CLARK, Breaking Defense
SPACE SYMPOSIUM: After years of dead-end efforts that cost more than $1 billion to build a more effective satellite tracking system, the US Air Force has begun using rapid software development techniques and cloud computing to build a completely new system.
The new approach is building a data cloud that may well expand beyond space to include all objects in the air and in orbit up to the Moon (cislunar for space aficionados). The head of Air Force acquisition, Will Roper, told me in an interview that this may include all data gathered for the Air Battle Management System, effectively giving the US military the ability to track and target every object flying and in orbit. That will include information from the new Space Fence, due to become operational this year, the ancient SPADOC system and other space situational awareness sources.
“We’ve taken a completely different acquisition approach now,” he said. This morning, Roper and Air Force Chief of Staff David Goldfein met to see a demonstration of this cloud-based data library. There were also space warfighters in the room. “The operators in the room said, even if it doesn’t get better than this, it’s already better than what we have today,” Roper told me.
(While Roper didn’t say so, my understanding is that this new system of cloud storage using apps will be part of Enterprise Space Battle Management Command & Control (ESBMC2), which will be used by the National Space Defense Center. The NSDC is the center where the military and Intelligence Community share SSA data and have command and control tools to fight a war that includes space.)
When Goldfein told Roper he should “replicate this kind of data library for the Advanced Battle Management System, my response to him was, why do a different one? Why not just keep expanding this one. It currently holds data from LEO to GEO. Why not include things you have in the air, or on the ground too?” And that’s how the Air Force would get to a single data system capable of tracking everything from the Moon on down.
But let’s not get too carried away. “We’ve got to earn our way to cislunar,” Roper said with a chuckle. “We’ve got to prove we can do LEO and GEO well.” Still, “this isn’t too far fetched.”
The old Space Situational Awareness (SSA) system, known as Joint Mission System, “was software intensive, (and) that hit the wall because 1970s era waterfall development just doesn’t scale to the software needs we have to write today,” Roper said when I asked if JMS was moribund.
So, forget JMS. The new system will be cloud-based and use a host of mission and planning apps, designed by any company ready to figure out new ways to manipulate the data and present it for use. This was sparked by a new colonel at Space and Missile Systems Center, Jennifer Krolikowski. She briefed Roper on a new software approach she thought could help.
“One of the great days for me in Air Force acquisition,” Roper said, “was the day that Col. K came and briefed me on her acquisition strategy, which was doing exactly that” — building software rapidly in increments. For the record, Krolikowski was sitting next to Roper when he said that. “We realized we have to do the same thing for JMS.”
She called it Kobayashi-Maru, which Roper noted “is the impossible training scenario that (Capt. James T.) Kirk hacks, so that it’s beatable.”
So far, Col. K (as she’s known, for obvious reasons) and her team have created two increments of the new system, one as a proof of concept and another, delivered Jan. 29, that allows satellite mission planning.
“All this information (from the cloud library) feeds in to my battle management, command and control tools,” she explained. That’s a very different tack from the one JMS took, which relied on the traditional building of a catalog of items in space and their positions at the time they were recorded. Those were then analyzed several times each day — once computing power grew enough. That process was slow and missed much and was certainly not geared toward warfighting in space.
Col. K has a “mandate is to provide space situational awareness, battle management and command and control tools for our warfighters, since they’ve been pivoting from watching what space is doing to warfighting.”
The mission apps, combined with machine learning and artificial intelligence, will completely change the warfighting dynamic in space. Using the current catalog system, with its regular conjunction analyses, is just too slow.
“A lot of stuff right now is very manually intensive,” Colonel K said, “so the time lines, if we wanted to provide an end-to-end kill chain, would be a little bit longer than we’d like.” But “when you automate a lot of those things you can get inside the enemy’s OODA loop.”
Now, how does industry play in this, you may wonder. “This is the first time I’m aware of where we’re treating the acquisition strategy of a program based mainly on the data it produces. We’re going to wrap an acquisition plan around the data,” Roper said. “This is a great area to bring in companies, so they can come in, look at the data and show us what they can do.” They then build the mission apps.
“The money we want you to make with us is not on archiving our data or storing our data or any of that we want you to make money by manipulating the data for warfighting lethality,” he said.
“I expect a lot of prizes and challenges and other interesting acquisition approaches that we’ve never been able to do at a large scale, because we’ve never been able to store data this way before,” Roper said.
One caveat about all this to bear in mind. “It’s not going to be one uber SA (situational awareness) program. It’s going to be lots of micro SA programs that share an architecture and interfaces,” the head of acquisition said.
The Air Force is still building a timeline for this. But it should move briskly along. Why? “This is being driven personally by Gen. (Jay) Raymond (head of Air Force Space Command and nominated to take over the revived US Space Command),” Roper said.