The goal would be to produce Toyota-like satellites, said Roper, “easy to build but reliable and hard working.”
WASHINGTON — The U.S. Air Force’s top acquisition official Will Roper wants future aircraft, missiles and satellites to be designed entirely in virtual environments using computer-generated models. That’s how commercial airplanes and automobiles are made, and it’s time for the military to get onboard, Roper said.
Air Force Secretary Barbara Barrett announced last week that the Air Force’s next pilot trainer airplane will be the first “e-aircraft” designed entirely in a digital format.
Roper said it’s too soon to reveal if or when the Space Force will have its first “e-satellite.”
“We do have two programs that are working on what could be the first e-sats for the department, potentially for the U.S. government,” Roper said Sept. 23 during a video conference with reporters.
Figuring out how the path forward for e-satellites “is going to take a little longer than where we are with aviation,” he said.
The two programs that will serve as test cases for digital design of satellites are classified so Roper could not disclose the names or the type of spacecraft. But he said that if these methods can be applied successfully, military satellites could end up costing dramatically less and get built in a fraction of the time it takes today.
Roper said a transition to digital engineering is going to require a major shift in the culture and the processes that have existed for decades in military space programs.
In fact, said Roper, many in the DoD acquisition world don’t really understand what digital design means.
“What I encounter with digital engineering is a lot of people saying we’re already doing it,” he said. But using computers to design something does not equate to digital engineering, Roper said. “How you use the tools matter.” Just because a sculpture is made with hammers and chisels doesn’t make it a masterpiece, he noted. “There’s a process for using the tools and there’s an art to using them.”
One reason e-satellites will not happen overnight is that they require changes across the entire supply chain. Each satellite program today has dozens of suppliers. “The supply base isn’t there, that’s the first thing that moves when you go to digital engineering.”
Even if virtual models existed to design, assemble and even operate satellites, “if my supply base doesn’t build parts that align with those models, especially their tolerances, then I actually don’t have a digital thread.”
Satellite program offices will have to build a digital thread “and then work with their vendors to meet the tolerances that their tools demand,” said Roper.
That would be a major departure from the traditional methods of building military satellites, Roper said. For digital engineering to work, the designs and technical requirements would have to be simplified.
“What I’m asking the two programs that are working on potential e-sats is to get rid of the hard, expensive or otherwise difficult things,” he said. “Let’s get rid of the high tolerance. cleanrooms, the expensive tooling, the highly experienced workforce, the rigging and harnessing, the things that make satellites expensive.”
The goal would be something akin to a Toyota-like satellite, he said, “easy to build but reliable and hard working in the space battlefield.”
Roper has argued that digitally engineered aircraft will save DoD billions of dollars in long term modernization and maintenance because the virtual models would be reused to inexpensively produce parts and components. Satellites don’t have a logistics tail like aircraft. “So the real return I’m looking forward in e-sats is a learning curve reduction, and the overall lowering of direct and indirect costs.”
Reducing the learning curve would bring cost down and shorten the development cycle, he said. “If we build the first satellite like we had built 20,” that would be a big deal.
The two classified programs that will test out digital engineering are managed by the Space Force’s Space and Missile Systems Center. Roper said he expects the experiment will take a couple of years. “I actually need to get my hands dirty in a specific program, because the devils will be in the details, but I’m very confident that something much better than the way we do it today will be achievable.”