The Search for Intelligence Autonomy

A science fiction world with autonomous aircraft — equipped with sophisticated sensors and onboard computers that gather and process data to make decisions and be able to perform without any human intervention — is about to become a reality with the help of Montclair State researchers.

“The Department of Defense wishes everything could be autonomous, from battle tanks to battleships — and even robot soldiers that can act like real human soldiers,” says Computer Science Professor Jing Peng. “But complete autonomy is not easy.” Peng and Linguistics Professor Anna Feldman recently received a Department of Defense-Air Force Office of Scientific Research grant for a project that addresses the challenges facing autonomy.‌

Aerial view of a dam.

The $145,170 award will support the purchase of powerful computer and sensor equipment that will help the researchers — and a team of graduate and undergraduate students — develop a system that gathers information in real time by detecting and characterizing short- and long-term events and activities. The system will also provide a platform for dynamic data processing, exploitation and management.

According to Peng, the new Department of Defense award also supports and expands Peng and Feldman’s ongoing National Science Foundation-funded efforts in the area of natural language and text processing — such as callouts or chat data — by analysts who conduct wide area video surveillance. “Callouts or chat data are text that we want to leverage to help us detect events and activities over live video surveillance streams,” he explains.

Video surveillance is a key component of United States Air Force and other intelligence analysis.

“Complete autonomy is not easy.” –Jing Peng

“When an unmanned aerial vehicle, or UAV, operates in a hostile environment, minimal communications with ground stations is a must,” says Peng. “This means we need the onboard capability to detect and characterize events and activities over live videos in such a way that only critical information — instead of every video frame — will be communicated to ground stations.”

By paving the way for autonomy, the team hopes to advance technology capable of gathering, analyzing, interpreting and protecting large amounts of diverse data to help with timely security decisions and policy making.