Rense.com



'Lifelog' Project Revived By
Defense Department

By Noah Shachtman
Wired News
9-13-4
 
It's been seven months since the Pentagon pulled the plug on LifeLog, its controversial project to archive almost everything about a person. But now, the Defense Department seems ready to revive large portions of the program under a new name.
 
Using a series of sensors embedded in a GI's gear, the Advanced Soldier Sensor Information System and Technology, or ASSIST, project aims to collect what a soldier sees, says and does in a combat zone -- and then to weave those events into digital memories, so commanders can have a better sense of how the fight unfolded.
 
That's similar to what planners at Darpa, the Pentagon's research arm, had in mind for LifeLog, its ambitious electronic diary effort. However, ASSIST's aspirations are more modest, its battlefield focus is clearer, and its privacy concerns are more manageable, military analysts and computer scientists say. All of that combines to give the project a better chance of taking off where LifeLog crashed.
 
"Welcome to the wacky ways of contracting at the Defense Department. If it doesn't fly the first time around, you can be sure it'll be back. And so it is," said Steven Aftergood, an analyst with the Federation of American Scientists. "This time around, though, the work has a slightly more plausible context. And more of an effort has been made to connect it to a military application."
 
Under the old Darpa program, every aspect of a life was fair game -- not just what was heard or said, but what e-mails were sent, what meals were bought and what TV shows were watched. Privacy advocates wondered how all that information was going to be used.
 
Darpa, in return, had a slew of answers. In a program overview, the agency suggested that the program could offer a way to create a computerized assistant for battlefield commanders. Darpa said the project could also provide a means to "support medical research and the early detection of an emerging epidemic."
 
That fuzziness set off alarm bells for civil liberties advocates, particularly because LifeLog came on the heels of Total Information Awareness, Darpa's unnervingly far-reaching effort to use ordinary citizens' records to profile potential terrorists. At first, the only people wearing the electronic diaries would have been those scientists developing LifeLog. Nevertheless, critics like The New York Times' William Safire were still concerned. LifeLog researchers might be comfortable recording their own actions. But what about the other people that the LifeLoggers are "looking at, listening to, sniffing or conspiring with to blow up the world?" Safire wondered.
 
ASSIST may run into similar hurdles.
 
Lee Tien, a staff attorney with the Electronic Frontier Foundation, notes that "I don't see anything about privacy in this (call for research) other than a note about privacy regulations in testing."
 
But the ASSIST project's explicit battlefield focus should ease some of the privacy concerns, said David Karger, an artificial intelligence specialist at MIT. LifeLog's tests were supposed to take place during researchers' trips to Washington, D.C. ASSIST's evaluations, on the other hand, are supposed to happen at one of the Army's urban warfare training centers, in Ft. Knox, Kentucky, or Ft. Benning, Georgia.
 
"By selecting a domain in which privacy is less important (than, say, keeping the soldier alive), I think it avoids some of the nasty social stumbling blocks that tripped up LifeLog," Karger wrote in an e-mail.
 
Darpa expects to hand out four- to eight-year-long early-development contracts for ASSIST, each for $4 million or less. The grants will be used to develop the hardware and algorithms needed for a system that "a dismounted soldier on a patrol" could use to record his time out in the field, according to a Darpa call for research. The GI would use a series of GPS locators, audio recorders, and video and still cameras to keep this diary. Those sensors could be activated "based on voluntary or involuntary physiological clues."
 
Once the patrol was done, ASSIST would automatically pick out "objects, scenes and activities" from the recordings, as well as identify key "events and states." Once all this processing was done, the data would be distilled into a "digital report that may support later patrols and mission planning," according to the agency.
 
The proposed project is similar to digital memory efforts going on at Microsoft and elsewhere. It would also be a big improvement over how so-called "after action reports" are currently handled, GlobalSecurity.org director John Pike noted. This could make ASSIST "very useful."
 
"In a dismounted battle, everybody has a different account of the action. Everyone's story is different. Everyone has a fragmentary insight into how the skirmish unfolded. That can make battles difficult to reconstruct, difficult to understand and difficult to learn from. It's hard to understand the enemy's tactics."
 
But Darpa believes ASSIST can do more. Eventually, it envisions using the diaries to map out cities and buildings and trade information in real time, not just after the fact.
 
Some researchers, however, are concerned that ASSIST may not go far enough. Darpa doesn't appear to be interested, for now, in including documents in the ASSIST logs.
 
That's "too bad," wrote Howard Shrobe, an MIT professor and former chief scientist at the Darpa division responsible for ASSIST, in an e-mail.
 
"The (armed) services, after all, are extremely document-oriented organizations. One goes into action with a set of orders, a situational assessment and a statement of the commander's intent; these are conveyed primarily as textual documents. If you're going to understand what happened as well as what was intended to happen, you'd have to understand these documents as well," he wrote. "Perhaps in future phases of the program this aspect of the problem might re-emerge."
 
To crunch all the information it receives, ASSIST will have to be smart and able to learn from the experiences its wearers feed it. Building these types of thinking machines has been the goal of Ronald Brachman since he took over Darpa's Information Processing Technology Office in 2002.
 
"It is the progressive improvement of the knowledge base of the system over time that we believe will best support soldiers on later missions," Brachman wrote in an e-mail. It will "allow them to understand what prior patrols saw and heard, and to recognize salient (and potentially life-threatening) changes in the situation when they go out on a mission."
 
© Copyright 2004, Lycos, Inc. All Rights Reserved.
 
http://www.wired.com/news/privacy/0,1848,64911,00.html?tw=wn_tophead_1


Disclaimer






MainPage
http://www.rense.com


This Site Served by TheHostPros