The Pentagon has given the Army the go-ahead to begin deploying a system that’s designed to take nine separate IT systems for intelligence gathering and analysis into a single cloud-based architecture.
The service has been developing the program, the Distributed Common Ground System-Army (DCGS-A), for a decade, and is a variant of the DCGS systems each of the military services are working on. The goal is to abandon the stovepiped systems the military has been using to collect and analyze intelligence and give tasking to its intelligence collectors — many of which use proprietary data formats and don’t interoperate easily — and replace them with a family of systems that speak a language that’s common across the intelligence community.
Frank Kendall, the Pentagon’s undersecretary for acquisition, technology and logistics, approved the full deployment decision for DCGS-A last Friday.
Many of the components of the system already are in place in Afghanistan, said Maj. Gen. Harold Greene, the Army’s deputy for acquisition and systems management.
The service deployed pieces of the system under special acquisition authorities for urgent needs in response to past intelligence shortcomings and were only available to units in immediate battle zones.
“So where we had this as a quick-reaction capability, we’ve now gotten the approval to field that capability across the entire Army intelligence enterprise,” Greene said. “It will allow us to standardize our training and standardize the program across the Army.”
No more proprietary data
Army officials say the elements of the system they’ve gotten the Pentagon’s blessing to deploy also are fully compliant with the overall intelligence community’s common architecture for IT systems.
Ditching proprietary data systems that couldn’t easily talk to one another was a major step forward for the Army, said Maj. Gen. Stephen Fogarty, the commander of the Army’s Intelligence and Security Command. “We’ve tested and fielded a lot of capabilities over the course of the war to answer a wide range of problems, and some have worked, and some have not,” he said. “I’m agnostic about the tools our analysts use, as long as they’re compliant [with the Intelligence Community’s standards], and completely interoperable. Any solution that uses a proprietary format that inhibits complete sharing of data, that’s a red line.”
Fogarty’s comments were an oblique reference to a system called Palantir, a data analytics platform developed by a private company of the same name.
DCGS-A received some unwanted national media attention over the past couple of years when service members wanted to use Palantir instead of the Army’s program of record, believing it performed some functions for intelligence analysts that DCGS-A didn’t.
Greene said the Army actually has deployed Palantir based on requests from commanders. Out of 13 requests, it fulfilled nine of them. But for now, the system doesn’t comply with the information sharing imperative for data interoperability.
The Army is hoping, however, that will change.
“There was probably a good reason units were requesting Palantir,” Greene said. “There were a number of comments regarding the ease of use. So as part of the process of having industry days, Palantir submitted a white paper.”
CRDA in place
The company was one of six submitters from an industry day earlier this year that the Army picked to enter into cooperative research and development agreements for more investigation into how its offerings would fit into DCGS-A instead of competing with it, Greene said.
“It’s an agreement where the two parties agree to share, and there’s got to be a benefit on both sides. We think they have some technologies that would be of value in DCGS-A,” he said. “The benefit to Palantir is that they’re now working in our labs, and they’re very familiar now with our infrastructure and our standards. Hopefully, they can become compliant with the intelligence community standards, because clearly, if there’s a capability there that’s best-of-breed and that our soldiers like, we want to go after that.” The version of DCGS-A that will start to deploy across the entire Army intelligence enterprise this year has one big gap — the parts of the new technology that are designed to handle top secret data aren’t ready for battlefield use and won’t be included in the first release. Instead, the Army will use the legacy systems it’s already been employing for top secret/sensitive compartmentalized information (TS/SCI).
That decision sprang from a critical report of DCGS-A earlier this year from the Pentagon’s director of operational test and evaluation. It turned out, Greene said, that while the system is designed to share information across agencies in compliance with the Intelligence Community’s policies for data standardization, it’s not ready for prime time when it comes to moving data across different levels of secrecy.
“What they found was there were some challenges with the workflows between the TS/SCI and the secret domains. They also found some reliability challenges,” he said. “So we’re going to deploy the other components of DCGS-A, minus the top-secret/sensitive compartmentalized information components.”
Next release a year away
Greene said the Army hopes to have the kinks in the top secret corners of the system worked out when the next release of DCGS-A arrives on scene in about a year. He expects those elements to be tested at the Army’s Network Integration Evaluation in the second half of 2013. In the meantime, legacy systems will do the job.
But Greene said the bulk of the heavy lifting those computers do in Afghanistan won’t be impacted, since the multinational coalition does most of its work on the Afghan Mission Network, which predominantly operates at the lower classification level of Secret.
“Part of this is indicative of what we’ve been doing for the past few years. So that’s where we put our focus,” he said.