MITRE | The MITRE Digest | Soldiers Grab More Intel with Web Media Technology

Home > News & Events > MITRE Publications > The MITRE Digest >

Soldiers Grab More Intel with Web Media Technology

April 2007

As Corporal Sandy Davis moves through the Iraqi village, he's constantly gathering human intelligence, or HUMINT, thanks to two cameras (one still, one video) plus digital writing paper. Cpl. Davis is learning about the people, their habits, and their relationships with each other. As he moves past the street vendors, he observes the flow of human interaction of every day life—what's normal and what's not.

Cpl. Davis snaps a picture of a strategically located office building with his digital camera, which has automatic image geo-coding. The camera's global positioning system (GPS) captures the building's latitude and longitude coordinates with an accuracy of 3 to 10 meters. When the photo of the building is printed, those coordinates are seen as part of the digital image.

Patrol track and associated video in a report editor. Both map and report media may be annotated.

A little later, Cpl. Davis uses his helmet-mounted video camera to capture images of a group of men that seem out of place. Then he uses the digital pen and writing paper to make notes about what he sees. The high-tech pen is equipped with a camera, infrared light, and computer chip. Every mark on the page is captured digitally and can be replicated as written on a computer screen.

Testing Geo-located Video

Paul House, a senior software engineer at MITRE in Harper's evaluation group, tested a video camera as he walked to work. House served eight years in the Marine Corps and was able to evaluate a geo-location system from the point of view of a military user. He found that the camera's accuracy would vary and later developed his own set of "Murphy's Laws."

"The geo-locating accuracy is very dependent on your GPS," says House. "When I was using a Garmin navigation device and it received only a couple of satellites, its accuracy was about plus or minus 40 feet. But when it received a dozen satellites, its accuracy was about plus or minus 6 ft or 9 ft. Every once in a while the GPS signal would get lost in the trees. When I looked at my route on the map afterwards, it showed that I wandered out in the middle of the street or walked through some buildings."

Some samples of House's set of "Murphy's Laws" regarding geo-located video:

The "+/- accuracy" is about 50% accurate.
Something will go wrong while copying files to the computer or downloading from devices.
You'll forget to take you lens cap off at least once.
The absolute most important aspect of this entire system is the GPS time stamp; without it, life isn't nearly as pleasant.

The data-capturing equipment in the above scenario could soon be used by soldiers to improve the collection, processing, and retrieving of intelligence information. In 2004, the Army realized that every soldier could be a sensor because they have the best local situational awareness. The Defense Advanced Research Projects Agency's (DARPA) Information Processing Technology Office (IPTO) set up a program to solicit technology proposals from commercial vendors.

"MITRE participated by acting as an independent technical evaluator during Phase I of the program, called Advanced Soldier Sensor Information Technology [ASSIST]," says Lisa Harper, a principal artificial intelligence engineer in MITRE's Command and Control Center. "We supported the IPTO with initial requirements development and evaluated early prototypes of tools. [See sidebar, "Testing Geo-located Video."] We also prototyped a reporting tool for searching and retrieving data."

Delivering a Bigger Picture

In the field, ASSIST would let soldiers upload their geo-coordinate tracking data and imagery into the Web-based tool so they can write textual reports that are annotated with their photos and videos. "We tested it with soldiers recently returned from Iraq. They said the new reporting tool saved a lot of time compared to manually keying reports into a word processing program," says Harper. "The photos and video are automatically geo-located on a map so that you can see where the media relates to a specific location."

Harper's group developed the prototype's basic application programming interface, its architecture, an initial set of interactive Web page views, and service connections to both photo and video data storage repositories. "We also developed a custom video player that enables users to easily interact with video using a map display, as well as a time slider that enables users to adjust when the video's time is off," says Harper.

After developing the prototype tool, it was transferred to SRA International for further development into an application called Tactical Ground Reporting Networking (TiGRNET), which specifically targets the needs of the U.S. Army. TiGRNET is a Java tool that SRA is building using its own proprietary components.

As a sociocultural sensor, ASSIST uses temporal pattern analysis of location-based multimedia content to answer questions such as: How has the population reacted to our presence and operations in Mosul over the past three months?

Customizing Broadens Usefulness

Harper's group is currently using its experience to develop a Web 2.0 service-oriented architecture, called TAGR, used for sharing location-based multimedia intelligence. Besides the military, potential users include law enforcement agencies and emergency service organizations.

"We believe there is a valuable opportunity to exploit the work we've done with rich Web 2.0 business models," says Harper. "Location-based organizational memory is impossible to maintain with each group creating ad hoc, isolated, and temporary systems. This is particularly the case for tactical organizations that make heavy use of ground-based imagery and other media."

TAGR is a platform designed for customization. The prototype is created from a mix of services such as Google Maps and Google Earth, open source Web development modules, and custom MITRE modules that are "mashable"—as in mashing together. "By developing TAGR, we will demonstrate a process for building loosely connected, adaptable, re-mixable, and easily maintainable systems," says Harper.

—by David A. Van Cleave

Related Information

Articles and News

Netted Sensors Issue of The Edge

Websites

Command and Control Center (C2C)

Page last updated: April 25, 2007 | Top of page

	Solutions That Make a Difference.® Copyright © 1997-2013, The MITRE Corporation. All rights reserved. MITRE is a registered trademark of The MITRE Corporation. Material on this site may be copied and distributed with permission only.
	Privacy Policy \| Contact Us