AFRL Develops Miniature Reconfigurable Sensor Processor for UAV Applications

  • Published
  • By Beverley Thompson
  • AFRL Sensors Directorate
AFRL developed a reconfigurable sensor processor suitable for unmanned air vehicle (UAV) applications. The unit's compact dimensions reflect the engineering effort to reduce not only the size of existing high-performance computing devices, but the associated power requirements and weight as well. AFRL's portable new real-time processor is powerful, too; based on a synthetic aperture radar (SAR) back-projection benchmark, it demonstrated a 36x improvement in processing time over a 2.8 GHz Intel® Xeon® microprocessor.

SRC Computers, Inc., devised the miniature unit's unique architecture, which pairs a general-purpose processor with the company's own MAPstation™--a reconfigurable processor based on field-programmable gate array technology. Once employed exclusively for traditional high-performance computing applications, the original MAPstation architecture has now undergone adaptation and ruggedization for use in embedded UAV applications, including sensor payloads.

The product improvements and applications realized as a result of this effort have led to several deployments of the SRC Computers system throughout both the Department of Defense and the private sector. Lockheed Martin selected SRC's MAPstation for use as the signal data processor in the Army's Tactical Reconnaissance and Counter-Concealment Enabled Radar (TRACER) program. TRACER will use a system containing four MAPstation processors to perform real-time SAR processing aboard the Warrior UAV.

Meanwhile, the Air Force Precision Image Tracking and Registration (PITrackR) program will use the portable MAPstation unit developed specifically for AFRL. General Electric Aerospace selected the miniature processor for PITrackR's use in detecting and tracking targets from an Arcturus T-16 small UAV. The 4-pound sensor processor will connect to an 11-megapixel camera through a CameraLink interface in order to perform onboard image processing and metadata correlation.