Kill Decision(42)
“And these are live images.”
“Live. Everything you see on the main screen is live, real world. But live imagery is the least of it.” The general moved over to a workstation manned by a uniformed JSOC first lieutenant wearing a headset—a strapping blond kid with clear skin and good posture. “Gartner, replay that truck sequence in sector H-Six we were looking at.”
The lieutenant immediately paused the imagery on his screen and tapped in some coordinates.
The general watched intently but spoke to Odin. “The critical difference of EITS over previous surveillance systems is that this high-resolution video imagery is retained over time in a data cloud, allowing analysts to ‘rewind’ the entire battle space—to see what might have taken place in a given locale over time.”
As Odin watched the nearby monitor, the image zoomed in to a tiny corner in the vast shantytown. People were walking past, but then they stopped in midstride. The video rapidly began to rewind, people and vehicles moving backward, until a faded red Toyota pickup moved into frame. The image then halted and began to play forward again, showing several armed men loading crates onto the truck bed.
Lieutenant Gartner was clearly used to doing demos with the general, because he was already doing what the general was about to ask.
“We can move in for a close-up of faces . . .”
The screen had already done so.
“. . . and we can even rotate the view.”
The image was already circling around to the other side of the men—not in a smooth pan, but in fifteen-degree leaps of POV.
Still, it was an impressive technical achievement. Odin remained emotionless. “How far back in time can you go?”
“As far back as we want to allocate storage space. We can even flag certain regions for long-term storage. Trouble spots.”
The image was already zooming out to the large city view. Live again.
“We use algorithms to parse human activity—tracking the pulse and character of a place. Automating what we call ‘pattern of life’ analysis. Compiling a fingerprint, a signature of a city’s normal routine. Airborne persistent video pattern-recognition systems will be big in this surveillance effort—Bayesian algorithmic models . . .”
The general was still talking as Odin watched a constellation of red glowing dots and squares superimposed on the vast city, like ants.
“This layer represents observable human activity. The dots are people, the squares vehicles. Over time the subsystem differentiates which part of the imagery is static city and which is dynamic human activity. But it goes further. Within that human activity layer, EITS begins to accumulate experience of the patterns of human living that represent a city’s background noise—its norm. What travel patterns are followed each day from location to location—with each dot being tracked representing a trip marker that’s added to the database. The totality of trips weaving a pattern of behavior. How consistent is this pattern? What portion of residents follow a routine, leaving and returning to the same places on a general schedule? Which portion of the population has no regular schedule? That lets us focus on areas of suspicious activity—a common point somewhere in the city where individuals who’d been present at earlier ‘trouble spots’ might later congregate, a place that might be the lair of an insurgent group—the sort of intel that your group would previously have had to obtain through HUMINT—can now be gleaned from observing the totality of human activity. Remembering it over time. Seeing everything. Forgetting nothing.”
Odin watched the companion imagery as Lieutenant Gartner played impressive visual accompaniment to the general’s pitch. Odin appeared deep in thought. “Our PIR usually involves locating a specific individual, and for that cell phone SIGINT suffices. Our knob turners can isolate known voice patterns, trace the—”
“You mean as long as you can run manned listening flights over the target area, and we already do that with unmanned airships that can stay aloft for weeks.” The general nudged Lieutenant Gartner aside and clicked through a few menus to bring up another information layer.
The screen suddenly flipped to an entirely new field of hundreds of thousands of clustered dots, moving through the city.
“Every cell phone’s IMEI and the base transceiver stations that serve them. This system simplifies eavesdropping. Just identify the phone you want”—he zoomed in and clicked on an ID number moving through central Brazzaville—“and you can record the subject’s communications.” The sound of foreign chatter came in over the speakers.
The general relinquished control to Gartner again and turned to face Odin. “Think about the combination of persistent telecom and video surveillance—being able to go back in time to see what happened on a street corner two months ago, before you even realized that someone was a person of interest.” The general gestured to an image of the huge city, clustered with dots. “This system displays the social map of an entire city from the communications and geolocation data of its citizens. . . .”