Still, senior Air Force officials on Monday asserted that they have addressed field testers' concerns as outlined in a final report, which they declined to release.
The chairman of the Joint Chiefs of Staff, Adm. Mike Mullen, is still "in favor of fielding this capability as soon as practicable, with the expectation that any technical issues would and could be adjusted and modified once it was in the field,'' said Capt. John Kirby, his spokesman.
The Gorgon Stare surveillance system consists of nine cameras attached to a pod that is designed to be carried on an unmanned aerial vehicle. It is supposed to provide images of a "city size" area.
But the draft report concluded that Gorgon Stare should not be launched until a number of issues were resolved.
"The Air Force can say anything it wants and talk itself blue in the face," said Winslow T. Wheeler, director of the Straus Military Reform Project at the Center for Defense Information, who circulated a copy of the report. "But none of the major problems have gone away."
The issues include deficient infrared performance, which would hinder the sensor's ability to capture images at night; inadequate resolution of images to track people on the ground in real time; a 12- to 18-second delay in receiving images on the ground; and "unpredictable" system reliability and lack of system documentation.
The critical report contradicts optimistic descriptions of the technology by current and former Air Force officials, who called Gorgon Stare "cutting-edge" technology in an article in The Washington Post.
In fact, some experts say, Gorgon Stare's wide-area surveillance technology is not as promising as other technologies, including one being fielded by the Army in Afghanistan.
Scientists with the Pentagon's Defense Advanced Research Projects Agency, for example, are developing a system dubbed Argus. That system is being designed to use algorithms to automatically track vehicles and even people over the whole area in view of the Argus camera.
By contrast, according to the Air Force memo, while Gorgon Stare imagery is "sufficient" to track vehicles, the real-time imagery is "generally not sufficient for tracking" people who "readily blend into the background clutter."
Moreover, the time it takes to download images to analysts after the aircraft lands is "unsuitable to conduct timely forensic analysis," the report said.
The final Jan. 11 report includes most of the same concerns as the draft but has reduced the number of corrective actions required before the system's deployment, according to an official who has seen the report and was not authorized to speak for attribution. The final report directs that the flaws should be fixed as soon as possible, the official said. Gorgon Stare is being developed by the Sierra Nevada Corp. in conjunction with the Air Force.
"Gorgon Stare will not be fielded until the theater commander accepts it," said Lt. Col. Richard Johnson, an Air Force spokesman. "The Air Force takes its responsibility seriously because lives depend on the quality of the intelligence products that are produced."
Wheeler, who has tracked waste and inefficiency in defense spending, said that beyond the $426 million already spent on Gorgon Stare and "closely related" Air Force research and development activities since 2009, the costs of the program will clearly grow.
"The saddest irony," he wrote in an e-mail, "is that even if Gorgon Stare were to be effective and reliable, it still would not help us win the war in Afghanistan, or anywhere else."