On a muggy day in DC, Elon Musk, the founder and CEO of SpaceX, stood in front of a gaggle of reporters and openly questioned the organization that could be his biggest customer.
“I don’t understand what’s taking so long,” Musk said at a June event when asked whether he expected the US Air Force to certify SpaceX’s Falcon 9 rocket for national security launches. “The Falcon 9 obviously works. It’s not as though the Air Force is changing the design of the rocket. They’re really just learning about it. That’s what the certification process is.
“So I don’t understand why it should take so long to learn about the rocket. That doesn’t make sense to me.”
It’s a familiar refrain to those who have been tracking the Air Force’s ongoing attempt at introducing competition into the national security launch market. The Falcon 9 has successfully launched, Musk’s supporters say, so what’s the hold up?
But according to Lt. Gen. Ellen Pawlikowski, the head of the Air Force’s Space and Missile Systems Center (SMC), the certification process is much more involved than simply reviewing documents.
Unquestionably, schematics and data from previous flights play a big part, but the service is also checking how conclusions in the data were reached, with a focus on processes and the back-end work that leads to a successful launch.
“How do you design? How do you manufacture? How are your operations conducted,” Pawlikowski said. “[We look at that so] we can have confidence that you have good processes in place and we can extrapolate the performance of your launches to cover the spectrum of launches that we need.”
While sympathetic to the slowness of the process, Pawlikowski said Musk should be aware of what he signed up for. She said Musk signed an agreement with “great, great detail” on what information SpaceX would have to provide.
“The document is actually a little over 200 pages long,” she said. “So there’s not any secrets about what the expectation is to be certified.”
The key to speeding up certification may lie in a decision made early in the process. At the start of certification, a company can choose from four options for the number of launches it will undertake: two, three, six or 14. The more launches, the less technical data it has to submit to the Air Force.
“There is a lot of information that we require to see when you only have three launches versus 14 launches,” Pawlikowski said. “If they had put forward that they wanted to go into that 14-launch column, then we would have required a lot less in-depth understanding of their processes, both manufacturing and design.”
When it signed its agreement in June to begin the certification process, SpaceX chose to go with three launches, likely as a cost-saving measure. After all, paperwork — even a mountain of paperwork — is cheaper than an actual launch. But the Air Force has found that working with technical information from a company such as SpaceX, with its own internal formats, schematics and data that doesn’t necessarily correspond with government standards the service would use, creates a certain amount of chaos.
When asked at a June 4 event about relations between the Air Force and SpaceX, Gwynne Shotwell, company president and COO, said the two sides are getting along, despite the frustration over how slow the process is. It was a conciliatory comment, one clearly designed not to further antagonize the people who may become SpaceX’s biggest customer.
But six days later, Musk didn’t bother to hide his frustration with service leadership and described why he decided to go ahead with a lawsuit against the service over its decision to award competitor United Launch Alliance (ULA) a block buy of launches.
“Our opposition is not with the Air Force, broadly. I think that should be emphasized,” Musk said. “We have a very close relationship with the Air Force, broadly. There are a few people at the senior level which — I just don’t understand where their opposition is coming from to competition. ... There’s a small number of people who are responsible for procurement decisions. We have an issue with them, not with the Air Force.”
Asked directly if he thought individuals in the Air Force were purposefully slowing down the certification process, Musk paused a full eight seconds before answering: “I think the jury’s out on this front. It doesn’t seem like it should take this long for experts in the launch business to understand how a rocket works, which is all the certification process is.”
Pawlikowski, who takes over the service’s top acquisition role at the Pentagon in July, reacted tactfully when told of those comments.
“Mr. Musk is a very successful entrepreneur,” she said. “I think what you’re seeing reflected in his comments is his approach to life. I don’t take it personally for the Air Force because I think that’s in [his] nature — he’s been very successful in a number of areas in taking on the culture.
“I understand why the process that we use is frustrating to him, but I can also tell you I’m not going to be the first SMC commander to lose a satellite on a launch,” she added. “I’ve got to balance the appreciation for how he’s been very successful at changing cultures while making sure I’m not sacrificing our high mission assurance that is our expectation as part of that culture change.”
Pawlikowski was more forceful in her response to Musk’s hints that service officials are purposefully slowing the program.
“I don’t know who he’s talking about, to be honest with you,” she said. “I can assure you there is no one out in Los Angeles that is anti-competition, because we’ve seen the benefit of it, or trying to slow roll the certification process. Whether he’s had some dialogue within the beltway, I’ve got no knowledge of it.”
Pawlikowski also pointed out that she has biweekly meetings with Shotwell, something Shotwell also commented on.
A clash between SpaceX’s Silicon Valley culture and that of the Pentagon was probably inevitable, said Brett Lambert, former deputy assistant secretary of defense for manufacturing and industrial base policy.
“These are two cultures coming to a new world at the same time,” Lambert said. “I think both sides need to adapt. I don’t think criticism on any one side is fair.”
Like Pawlikowski, Lambert doesn’t believe anyone in the Air Force is purposefully throwing up road blocks in SpaceX’s way.
“The Air Force is such a huge organization, it’s like asking which state in the union is putting up resistance to one thing. It’s too big to group in those terms.
“I think it would be fair to say there are people, not just in the Air Force but in any Pentagon office, who are more forward leaning than others,” he said, before calling Pawlikowski one of the “more forward leaning, creative and open-minded” officers in the service.
Complicating the process is the presence of competitor ULA, which recently hired public relations firm Shockey Scofield Solutions to help defend itself against the PR broadside issued by SpaceX. As a customer for both companies, the Air Force finds itself caught in the middle.
“Both sides are engaged in a very public PR battle,” said Brian Weeden of the Secure World Foundation. “It’s very hard to cut through all that PR and spin to really figure out what the hard facts are.”
The second time through an acquisition process is always easier, or so the prevailing theory goes. Does that mean that when SpaceX begins the process of certification on its Falcon Heavy (a process expected to begin late this summer) certification should take less time and money?
“Whomever follows SpaceX will be the beneficiary of both their efforts and the Air Force efforts,” Lambert said. “It will be much easier the next time.”
Weeden isn’t so sure. “I think the likelihood of speeding up the process is probably pretty low, given both the service’s lack of experience in doing it and the culture of risk aversion.”
Pawlikowski agrees there will be some lessons learned, again highlighting the question of how many launches to do for a new entrant, and said she expects to find some limited efficiencies. But she warned against expecting a major change in either time or cost for the next entrant.
SMC estimates that certification for the Falcon 9 will end up costing the Air Force around $100 million. Asked whether there is a way to drive that down, Pawlikowski paused and said, “I don’t think so. Obviously we can become more efficient, but I’m not sure that I can drive it down significantly”
That’s because so much of the cost is wrapped up in aerospace engineering personnel who have to review the documents from the new entrant. Those personnel largely come from the Aerospace Corporation, a federally funded research and development center that Weeden calls “the brains of the Air Force when it comes to really difficult engineering and systems operations problems.
“Part of the reason it’s costing so much is because the Air Force has to basically pay for a whole mess of aerospace contractors to spend a lot of hours and a lot of time to pore through data, do the calculations and do the analysis,” Weeden said.
By Pentagon standards, $100 million isn’t a huge amount, but for SMC — which saw its budget cut from $8 billion in 2012 to $5.6 billion in 2014 — it’s not pocket change. However, Pawlikowski expressed confidence that savings would quickly cover the cost.
Just the threat of competition from SpaceX is driving down the prices offered by ULA, she said, and SpaceX has already offered relatively low cost for launches.
“Between driving down ULA’s prices through competition and leveraging the prices that SpaceX is advertising, I think you’ll see that we will recover that $100 million pretty quickly.”
That is, if history doesn’t repeat itself. The Evolved Expendable Launch Vehicle (EELV) program under which military launches operate was designed to go hand-in-hand with a robust commercial launch market. That market never materialized, leading to rising costs and the eventual decision to merge Lockheed and Boeing, the two EELV entrants, into ULA.
“The whole reason for doing the certification of new entrants is seeking to see if we can leverage the commercial launch market. We’re not doing this so we can have another vendor or two more vendors that are 100 percent reliant on national space missions,” Pawlikowski said. “If we do that, we won’t save any money.” ■