An internal Army memo is calling on officers to fight legislation from a Marine veteran in Congress who wants to mandate fixes to a complex intelligence computing network panned by soldiers at war.
The congressman in question, Rep. Duncan Hunter, says that the March 22 email is essentially encouraging lobbying, which is outlawed for the military.
Meanwhile, a new test report on the war-deployed network, the Distributed Common Ground System, found that the latest upgrade, known as “Increment 1, Release 2,” is “operationally effective.” But the report is a mixed bag. It quotes battalion commanders as saying they did not find it “very helpful” during battle. They said they used “pencil and paper” to track the fight and later added the details to DCGS.
DCGS also provided inaccurate weather forecasts, according to a copy of the 2016 report obtained by The Washington Times, and it requires soldiers to undergo constant training.
The email from the Army National Guard intelligence directorate, or G2, which works closely with the active Army headquarters, also criticizes The Washington Times for its series of stories on DCGS and its spotty performance in the Afghanistan war.
It asserts the newspaper is a “lobby” for the Pentagon to buy an off-the-shelf data processor called Palantir, developed by a Palo Alto, California, technology firm. A functioning DCGS is crucial to fighting hard-to-find insurgents and terrorists. Analysts need to retrieve and study a wide range of reports to help commanders plan operations.
Palantir, a relatively simple network, has won praise from warfighters and is in use in Iraq. Army special forces specifically requested Palantir as they deployed back to Iraq in 2014 in the new war against the Islamic State army.
The National Guard email was sent to a long list of officers informing them that the Army opposes draft legislation from Mr. Hunter, California Republican and a member of the House Armed Services Committee.
His bill would order the secretary of the Army to restructure DCGS. It would also tell the Army to buy functioning, off-the-shelf software, rather than have Army contractors develop new programs.
Previously, the Senate Armed Services Committee adopted language encouraging the Army to buy commercially available computers while DCGS received fixes.
The Army March 22 memo states that buying ready-made software would not work. “Delivering the same capability of DCGS-A with COTS [commercial off the shelf] creates compatibility and lifecycle/software update conflicts,” it says.
“We are asking that you approach your federal House of Representative and Senate leadership to provide them your input and advice as military intelligence senior leaders in the Army National Guard,” the memo says, according to a copy obtained by The Washington Times.
Mr. Hunter, who served in the Afghanistan and Iraq wars, has read a steady stream of leaked memos from commanders and their soldiers in the field who complained that DCGS is slow, too complex and prone to crashes. In some cases, units in Afghanistan parked DCGS hardware in a corner and used commercially available computers.
The National Guard email aligns with the Army’s fierce loyalty to DCGS. The effort has included blocking war-bound soldiers’ attempts to win Pentagon emergency approval to buy Palantir, The Washington Times has reported.
The Associated Press reported in March 2015 that, “Special operations troops heading to war zones are asking for commercial intelligence analysis software they say will help their missions. But their requests are languishing, and they are being ordered to use a flawed, in-house system [DCGS] preferred by the Pentagon, according to government records and interviews.”
The National Guard email also contains a reprint of a Washington Times story that quoted retired Lt. Gen. Michael Flynn, former director of the Defense Intelligence Agency, as saying DCGS has failed to live up to its promises and should be revamped and renamed.
In a statement, Mr. Hunter’s office called the National Guard email “more evidence” of a long-standing campaign “within the Army to undercut attempts by Representative Hunter and others to turn DCGS into an effective system.”
“By no means should anyone in the Army be encouraging others in the organization to lobby Congress on an issue, which is prohibited,” the statement said, accusing the author of “an agenda [he] has no business pursuing.”
Mr. Hunter’s office also accused the National Guard email of “inaccuracies and misleading statements” about the Times article.
“General Flynn’s comments came from experience and there’s tremendous value in his insight. He’s not on Palantir’s payroll nor has he ever been. And he’s not the first high ranking officer to say DCGS is a train wreck. He’s just the first to say it publicly,” Mr. Hunter’s office said.
Asked to respond, the Army issued a statement to The Times, referring to a new report on 2015 weapons testing from the Pentagon’s Director of Operational Test and Evaluation.
“In December 2015, Distributed Common Ground System — Army (DCGS-A) received positive feedback from both the Army Test & Evaluation Command (ATEC) and the Director, Operational Test and Evaluation (DOT&E) after robust operational testing of the program. Report results state that DCGS-A addressed a number of critical issues identified in previous testing, including increased intelligence clarity and improved system reliability and stability,” the statement said.
“Working in collaboration with major Army commands and units, DCGS-A Program Management Office is currently fielding the most user-friendly, effective, and interoperable version to date. There is approximately 80 percent commercial software being used in the current version. The Army looks forward to DCGS-A Increment 2 [the next version] that maximizes competition among commercial vendors and harvests the best capabilities to support the timely, secure and effective exchange of intelligence information.”
The 2015 report on an upgraded model marked an improvement compared with DCGS units previously sent to war. The 2012 tester’s report said that version was “not effective, not suitable and not survivable.”
The 2015 report says that the latest version, “Release 2,” is “operationally effective.” It was deployed in February.
Yet, it continued to receive low marks for ease-of-use, a consistent complaint from soldiers at war. The report said that “users rated the usability low-marginal” and the system is still “not survivable against cybersecurity threats” because of an Army network problem. This presumedly means enemy hackers can break into the network.
The Times obtained a copy of a subsequent, more detailed, Pentagon test report dated January 2016.
The report by top tester J. Michael Gilmore, states that DCGS still has “systematic shortfalls with data collection [and] analysis.”
But overall, he wrote, “The test unit successfully received, processed, exploited and disseminated intelligence data with DSCG-A. The unit provided actionable intelligence to commanders, enabling them to make timely decisions.”
The testing involved creating “vignettes,” such as locating a terrorist leader or a bomb factory or disrupting a planned suicide bomber attack.
The report concluded that Release 2 is “operationally suitable” but with a caveat. This happens only if the Army “intensively trains DCGS-A users and provides continued refresher training to units in garrison. DCGS-A is a complex system and the skills required to use it are perishable. Partly because of the complexity of the system, the analysts stated they cannot maintain high level of skills without constantly using the system.”
• Soldiers complained they had trouble finding “relevant results” when searching DCGS’s Integration Backbone for data from all sources.
• In changing shifts, the incoming soldiers could not work on the previous shift’s intelligence product. Instead they had to copy and post it and then start a new product. “This wastes time and results in many versions of the same DCGS-A products,” the test report said.
• “Comments from battalion commanders and staff indicated they did not consider DCGS-A to be very helpful for the fight on the ground. They stated that once the battle starts, it is very difficult to update DCGS-A while also tracking the battle. As a workaround, some battalion analysts resorted to tracking the battle using pencil and paper and updated DCGS databases once the battle was over.”
• A weather staff officer said DCGS produced “inaccurate weather predictions.” Instead, the officer used the Air Force weather website. It turned out, DCGS cannot accept the Air Force’s three-letter location codes; It was designed to only accept four-letter codes.
• At the battalion level, the command closest to the battle, analysts “expressed frustration with the amount of time they had to spend working on the database. They complained it can take hours to synchronize databases.”