This post is also available in: heעברית (Hebrew)

The large amount of data streamed from a wide variety of sensors has been a challenge to the data fusion research community.  Weapons developers in the US Pentagon are analyzing emerging airborne sensor technology designed to fuse and organize data from multiple sources to provide faster and more accurate intelligence information.

According to defensesystems.com, Lockheed Martin is testing computer processing improvements to an Airborne Multi-INT Lab (AML) technology which integrates electro-optical sensors with synthetic aperture radar and other kinds of electronic intelligence gathering mechanisms to expedite the delivery of decision-quality intelligence. The AML system is currently configured to fly from a Gulfstream business jet; however, in the future, Lockheed plans to work closely with the military services to equip drones and fixed-wing aircraft with the technology.

New computer algorithms can transform raw data into intelligence by connecting disparate types of sensor data. Updates were made to AML’s onboard processing capability to enable this signal intelligence (SIGINT) and communications intelligence (COMINT) technology.

“Think of COMINT capabilities for direction-finding and geolocation and audio,” said Joe Suhr, chief engineer, research and development lead at Lockheed Martin. “The AML now has an autonomous sensor control mode that can coordinate operations between the testbed’s various onboard sensors,” Suhr explained.

The continued emergence of artificial intelligence (AI) systems allow computers to perform procedural functions without needing human intervention at each stage. Developers explain that human cognition is still needed to solve problems and, at times, respond to fast-developing unforeseen circumstances, but machines are now able to use algorithms and processing power to condense, align, organize and display combat-relevant sensor data by themselves.

The enhancement of real-time video coming straight off the sensor allows more efficient consumption of intelligence information. The direct video connectivity permits ISR video to arrive in the aircraft directly without having to first transmit through a ground station, Suhr explained. The technology not only decreases time of video feed delivery, but also better organizes incoming ISR information, allowing for combined data consumption and improved sensor interoperability.

The interoperable sensors are engineered for cross-cueing to fuse and interpret data streams. Antennas on the aircraft operate in a range of frequencies, Suhr explained.

This kind of integration among otherwise disconnected sensors contains a clear parallel to the technology built into the Pentagon’s F-35 Joint Strike Fighter. Described by operators and developers as sensor fusion, computer algorithms allow for information from a variety of sensors to be organized or fused onto a single screen for pilots to view. Sensor fusion for the F-35 also represents progress in the area of AI as it integrates ISR, targeting, navigation and radar warning receiver information autonomously.