Research arm for Intel agencies looking for nontraditional sensors.

• Author: Magnuson, Stew

* The Intelligence Advanced Research Project Activity is funding long-term research for several sensors, although not necessarily the kinds that are installed in a ball underneath an unmanned aerial vehicle.

"Sensors produce data and people make decisions. It has always been that way," said Peter Highnam, the agency's director. LARPA stresses the human factor in almost everything it does. "It's all about people," he said of the programs the agency is funding.

It has only existed since 2006, and is modeled on the more famous Defense Advanced Research Projects Agency, DARPA.

It "invests in high-risk, high-payoff research programs that have the potential to provide the United States with an overwhelming intelligence advantage over future adversaries," the agency's website said. It funds three to five-year projects, and is not interested in "low-hanging fruit," it added.

Highnam said IARPA serves 16 agencies, about half of which are in the Defense Department. The rest are in the intelligence community, along with some surprising customers such as the Treasury and State Departments.

Those two departments have just as hard and just as interesting problems as anybody else, he said. They had been "crying out for technology help to come in."

Most of the agency's customers don't actually control the sensors that provide them with data, he noted.

"What happens when we don't control the sensors, but we have to make use of the data that comes from the technology?" he asked at the Defense Security and Sensing Conference in Baltimore.

Sorting through and making sense of the glut of data in order to "make tough decisions" has been a widely reported problem of late, but it has always been an issue in the intelligence community, he said.

"People use the data that comes from these sensors, and frequently people are the sensors," he added.

One example of "people as sensors" is the TRUST program, or tools for recognizing useful signals of trustworthiness.

"If I meet you for the first time, how do I know I can trust you? This isn't a lie detector test. This is how do I know I can trust, begin to work with you and do something productive?"

Research shows there are physiological indicators that are generated within that give a person a good initial understanding of whether someone is trustworthy. But then cultural components override these signals and that insight is lost, Highnam said.

The question is whether an intelligence officer can become a sensor and pick...