Sensors are ubiquitous, providing autonomous and manned systems in the air, ground and sea the ability to “understand” their surroundings and relay that information to the end users and decision makers. The computational complexity associated with processing data from these sensors has grown to extraordinary levels as a result of new generations of high capability sensors and the fusion of multiple sensors together to provide a single operating picture of the surroundings. QEDLabs and Black River Systems Company propose the development of a novel high performance embedded computing (HPEC) platform incorporating a heterogeneous processing architecture of CPUs, GPUs, and FPGAs with complex algorithmic sensor processing software. The proposed system aims to improve fidelity, quality, and response rate for real-time advanced sensor processing algorithms in Radio Frequency (RF), Synthetic Aperture Radar (SAR), broad area Infrared (IR) and Electro-Optic (EO) Imaging, allowing tracking, target detection, and identification of real-time and persistent surveillance data.
Benefit: When commercialized, the platform will enable more advanced sensor processing algorithms in a smaller size, weight, and power envelope than available today. This will lead to better decision making and in the end save lives. We expect our commercial system to be deployed in mid- to high-end UAV systems either as technology insertion are as built-in capability for new systems. We also expect variants to be incorporated into smaller UAVs, ground, underwater, and surface unmanned vehicles, and in manned systems.
Keywords: Hpec, Sensors, Algorithms, Persistent Surveillance, Hyperspectral Imaging