How DARPA's Mosaic program could disrupt federal contracting

DARPA’s Mosaic Warfare project is taking on the vulnerabilities of large, complex weapons systems and in so doing may disrupt traditional ways of procurement.

This is the Defense Advanced Research Projects Agency's newest attempt to create a process for pulling together different capabilities into weapons packages that overwhelm adversaries. Think of each system as a different tile in a mosaic.

But don’t fall into the trap of thinking of it as a so-called "systems of systems," said Timothy Grayson, director of DARPA’s strategic technology office. He was a speaker at the fourth annual Doing Business with DOD and the Intelligence Community event produced by Jennifer Schaus and Associates and the Virginia Procurement Technical Assistance Center at George Mason University.

The U.S. military needs a different way to field and deploy warfighting capabilities as way of countering the growing military strength of countries such as China and Russia.

“It isn’t just that they are building new weapons systems, it is the speed with which they are building them,” Grayson said.

The U.S. might be technically superior right now but the curve of new development being demonstrated by China and Russia is a risk, he said.

“One of the thoughts of staying ahead is to not rely on the next big, monolithic program of record,” Grayson said.

Parts of DOD are pushing to create a system of systems but there is a lot of risk with the approach and it has failed in the past. Grayson cited the Future Combat System effort as an example.

“DOD likes to do things very requirements driven and there is a danger of moving very aggressively to a system of systems mindset where the very architecture becomes a monolith itself,” he said.

Monoliths just to be clear are big, complex, expensive and inflexible -- all negatives in a rapidly changing threat environment. What is needed in today’s environment is adaptability. That means pushing more of the decision making and development work out to the operational edges, Grayson said.

There are examples of that happening now with the Air Force’s software engineering groups that are bringing new capabilities through operational sustainment work on weapons systems.

“Right now there aren’t a lot of companies that can provide that kind of service,” Grayson said to the audience in a not-so subtle hint at a possible business opportunity.

The idea is to add new capabilities quickly. To do this, DOD needs to operate like a commercial entity and move away from cost-reimbursable contracting. Cost-reimbursement puts the risk and responsibility on the government, which pushes requirements down to the contractors. There is no incentive for the contractor to be innovative, Grayson said.

Moving away from complexity is another goal. Too often systems are developed with a 30-year horizon in mind so they end up having what Grayson called a “Swiss-army knife-set of capabilities with hundreds of functions” because they are guessing what the future needs will be.

But if new capabilities can be added quickly, complexity can be replaced by speed, he said.

Through Mosaic, DARPA is looking at three things that will impact procurement: Planning, Interoperability and Execution.

With planning, the issue is how do you decide what you need? What kind of planning aids and decision making capabilities are there?

Interoperability touches on how to connect with legacy systems and use virtualization so new systems and legacy systems can work together. It also entails greater use of machine learning.

For Execution, the focus is on how do you humans and machines interact. What can be automated so that humans focus more on battle management issues?

To get there, DOD needs to buy new capabilities as a service. Grayson cited two examples where this has been done: the EnhancedView program for the National Geospatial-Intelligence Agency to buy additional imagery from commercial satellite providers and the CIA’s cloud infrastructure contract with Amazon Web Services.

In both cases, the private sector made the investment in hardware and software and the government bought the capabilities as a service, Grayson said.

“It is an interesting model. It shifts risk to the commercial players but it unleashes innovation because you don’t have a program office telling you how to do it,” he said.

While he didn’t mention the JEDI cloud infrastructure contract, it was developed based on lessons learned from the CIA cloud initiative.

Approaches such as EnhancedView and cloud initiatives need to be more broadly adopted, he said.

“You design the capabilities and productize it and we’ll buy it back as a service,” Grayson said.