Inside SAIC's 'Innovation Factory' and the market's mindset change

The opening of Science Applications International Corp.’s “Innovation Factory” in October of last year and subsequent work to further stand up the initiative helps illustrate a shift by SAIC and other government IT contractors toward a new model.

In are more incremental and faster deliveries of technologies as industry and government agencies increasingly look toward an approach of agile development and deployment.

Traditional multi-year development programs often dubbed the “waterfall model” are not necessarily going away, but agencies are taking the quicker cycles of innovation into account in how they structure IT contracts that may start out small.

That shift in agency thinking certainly puts more onus and stress on the system of companies like SAIC that traditionally are wired to do business in a certain way and particularly emphasizes a change in mindset, SAIC Chief Technology Officer Charles Onstott said.

“It really requires us to rethink the processes that we use in how we deliver, it makes us rethink how we measure success and then in how we think about what we will bid,” Onstott told me. “Because some of these things are going to be relatively small engagements that in the past may not have been that interesting, but now they suddenly are interesting because they might lead to a big solution later on even though you start out kind of small.”

There is also the physical change. SAIC renovated part of a floor in its Reston, Virginia, corporate headquarters to house the Innovation Factory that has conference rooms of teams working on different agile development projects.

Collaborations with technology companies are also a part of SAIC’s approach to the factory, which includes Red Hat as a partner. SAIC also has office space in Austin, Texas, to connect with technology startups there and further help cultivate an ecosystem of partners with innovative offerings.

One aspect of the change in mindset is the deadlines the delivery teams face, sometimes within weeks or days to develop a prototype product ready for testing.

Take for instance the Air Force’s Pilot Training Next initiative, through which the service branch took an experimental approach aimed toward speeding up how quickly pilots can shift from simulation and into more realistic training programs.

SAIC collaborated with the Air Force to create a new simulator but in a way far different from the usual multiyear approach with much testing and design time built in, plus the matter of identifying requirements.

Typically, pilots would need 15 months to be ready for flight school. But this program had them ready to go within six months.

A solution for Pilot Training Next was developed and deployed within 45 days, Onstott said.

“We actually built a prototype solution within a couple days and then put pilots in it and said ‘Okay what else does it need to do that it’s not doing?,” he added.

Onstott’s description of how SAIC built the simulator also helps illustrate the larger shift in delivery times.

“This is all using off-the-shelf products… stuff we bought literally on a credit card and assembled together so it’s like a VR helmet and a gaming computer… basic stuff you can get,” he told me. “It wasn’t like we had to go custom design and manufacture to some special specification.”

Pilots that trained in the simulator gave constant feedback on what improvements they might make to help create a more continuous loop of communication. That quicker and more continuous feedback loop is another aspect of the mindset change for agile delivery, Onstott said.

“It’s not the old school ‘We start with requirements, we go through design, we then create a solution and then a year later we have something,’” Onstott told me.

This alternative method as “We create a prototype, we get people actively involved in the evolution of that prototype, and then it’s actually used,” he added.

Pilot Training Next also helps show how government agencies and the Defense Department especially are rethinking what their mindsets should be surrounding the element of risk. This program was labeled as experimental at the outset to take more risk and acknowledge it may not have worked.

But there is a paradox to that changing mindset toward risk, according to Onstott.

“The government is really mitigating their risk through embracing uncertainty,” he said. “Normally we think of mitigating risk as really increasing our certainty and line-of-sight on what the outcome is going to be.”

Rapid acquisitions like Other Transaction Authorities, plus data or other technology challenges agencies issue are helping shape that picture. In the Pilot Training Next example, that program was broken up into three-week sprints to check progress and bring in new solutions as needed.

“What we’re seeing in these new models is that… they have a good idea of what the outcome is going to be but they don’t know how to get there exactly. These incremental delivery models actually de-risk that situation,” he added.

“Agile in general is the main ‘secret sauce’ around how you embrace more uncertainty and more innovation but manage your risk at the same time. Now you’re not waiting two years to find out if this thing is going to work or not, you find out within a few weeks whether you’re on the right track or not.”