Air Force confident it can build satellites faster. A bigger concern: software

by
Only a few weeks into the job, Roper is overseeing wholesale changes in how the Air Force buys space technology and manages programs.

WASHINGTON — Space modernization is “one of the cool parts of this job,” said Will Roper, assistant secretary of the Air Force for acquisition, technology and logistics.

Only a few weeks into the job, Roper is overseeing wholesale changes in how the Air Force buys space technology and manages programs. The battle cry is to move fast, especially with satellites. The Pentagon is trying to replace traditional constellations with new systems that are more resilient to jamming and cheaper so they can be deployed in large numbers.

“I’m excited to be a part of space modernization,” Roper told reporters on Friday at the Pentagon.

“Now that we’re thinking of peer competition, what’s great is that we can take a lot of things we learned in the air domain and think about their analogues in the space domain,” he said. “I’m very happy to see a lot of creative thinking in space right now.”

The Air Force is trying to show it can innovate military space systems as lawmakers continue to push for change in even more aggressive ways. In its markup of the National Defense Authorization Act for Fiscal Year 2019, the House Armed Services Strategic Forces Subcommittee directs the deputy secretary of defense to “develop a plan to establish a separate alternative acquisition system for defense space acquisitions, including with respect to space vehicles, ground segments and terminals.”

Roper said he has not had “any engagement with Congress on the pros and cons of a separate force.” So far, “I’ve been focused on understanding where space acquisition is.”

Missile warning satellites
The plan over the next several months is to complete a reorganization of the Air Force Space and Missile Systems Center, in Los Angeles. The center manages a $6 billion portfolio including the majority of the Air Force’s space programs. The immediate test for the reorganized SMC will be to start developing a constellation of missile warning and surveillance satellites to replace the Space Based Infrared System. The Air Force bought six SBIRS satellites and decided it needs a different system in the future. SBIRS is the poster child for “exquisite” satellites that make attractive targets because of their operational and financial value.

The future missile warning system will be a “pacesetter” for learning to speed up traditional acquisitions, said Roper. Being a pacesetter “is more than just building a prototype or a low cost system.” The idea is to “drill down” into the reasons why programs always take longer than predicted, he said.

The goal for the SBIRS replacement is five years. Roper hopes to “switch the mindset” of procurement managers so they do a better job balancing the need to deliver on time with a “reasonable amount of experimentation and prototyping.”

Developing, producing and launching into orbit a new constellation in five years is “aggressive,” said Roper. “Five to six years is a gold medal.” Whether it’s five or six years, the idea is to start changing the thinking “so program managers can take advantage of discovery in getting things right but can hedge their bets in case something goes wrong.”

“I’m confident as we learn things we’ll adjust our vision,” Roper said. The Air Force decided that five years is a reasonable goal to get the workforce motivated. “Until you write down the best you can do and go for it you’ll never achieve,” he said. “You’re not going to take a nine-year schedule and it magically becomes five.”

The “litmus test” for the SBIRS replacement program will be how leaders cope with setbacks. “If a slip happens and we deliver in six years, do we view that as a failure? If we do, then this ’embracing failure and going fast’ experiment isn’t going to work very well,” he said. ‘We have to be happy with results that are statistically better than the norm.”

If the new constellation gets off the ground, makes it through the hurdles and “isn’t punished” when discoveries are made that cause delays, “other programs will follow,” said Roper. He wants this to work. “It’s a very important program for the department. And it’s important to me personally because I believe other programs are going to be watching it and taking notes.”

Problems buying software
Speaking broadly about what ails military procurement programs, Roper said he is generally satisfied with the performance of Air Force programs — but only on the hardware side.

“Our big issue is software,” he said. “Almost every software intensive program is over budget and behind schedule.”

The Defense Science Board probed the Pentagon’s troubled software acquisitions and offered some proposed fixes in a recent report. The panel of advisers said the Pentagon has to get on board with commercial “iterative software development” practices where engineers make rapid changes, ask for user feedback and adjust the software for the next increment. Most defense programs use traditional “waterfall”  development  — sequentially going from requirement to software development and then testing. The waterfall approach was mostly abandoned by commercial companies in favor of agile development.

Some of the biggest software headaches for the Pentagon are the F-35 Joint Strike Fighter and the next-generation operational control system for GPS satellites, known as OCX. To help turn things around, the undersecretary of defense for acquisitions, Ellen Lord, recently hired a noted engineer and technology leader Jeff Boleng as special assistant for software acquisition.

Roper said he is a major proponent of agile software methods.”I have to get the Air Force to where we can do agile software development.” The Joint Strike Fighter will be especially challenging, he said. “It’s a software program at this point. We have to show we can pivot, that we can do software drops like commercial, every couple of months or every couple of weeks.”