Army innovation lab tests coders on the battlefield

The Army's newly launched Army Applications Laboratory is working to expand ability for soldiers to code in the field to repair communications systems.

Army Futures Command's innovation lab is concocting solutions that range from putting coders on the battlefield to using machine learning to automate the budget decision-making process.

Gen. John Murray, Army Futures Commander, told reporters following a House Armed Services Subcommittee hearing on Army modernization that the service's newly launched innovation lab, Army Applications Laboratory, is working to expand ability for soldiers to code in the field to repair communications systems.

"We actually forward deployed some coders to work with units on problems as they were discovering [them], the ability to recode mission command software to account for some things that weren't readily apparent to the operational force," Murray said.

The Army is looking to expand that capability, possibly in the multidomain task force, which has been focusing on interoperability challenges. Murray said that having soldiers who can code on the fly is an important skill the Army needs for the future of warfare.

"I do think that's a skill set that the Army is going to have to adopt: some really smart kids who can code so a commander can find out a problem -- C2, command and control systems aren't working right -- go get his four hours of sleep, which is mandatory every night, come back and in the meantime, we've re-coded the system to account for the problems that we're having."

The Army Applications Lab is also working on a solution that would use machine learning to automate budget decisions. The Army completed a nearly year-long evaluation of its programs through a zero-based approach, justifying every dollar spent. Murray said the lab is looking at algorithmic solutions that would cut man-hours needed and evaluate risk to the operational force and industrial base when procurement levels are reduced.

"It's not necessarily the decision-making, but how that is presented in a more automated way. Right now, that process was a stack of Powerpoint slides about this high," Murray said gesturing his hands a few feet apart.

Using machine learning would allow Army leaders to "see the impacts of the decisions in real time as opposed to somebody going back and working on it for six or seven hours and say this is the impact of the [budgeting] decisions made," he explained.

"A more automated way of not making the decisions but really providing the information to the decision maker."