Agile Condor

The Agile Condor is a podded edge-AI processing unit designed to run computer-vision targeting algorithms directly on board the MQ-9 Remotely Piloted Aircraft (RPA). Developed by SRC Inc. and integrated onto the airframe by General Atomics Aeronautical Systems (GA-ASI), the system was introduced in 2020. Its primary function is to provide on-board high-speed computer processing coupled with machine learning algorithms to detect, correlate, identify, and track targets of interest without relying on a continuous data link to ground stations.

The Agile Condor operates by leveraging the MQ-9’s existing Electro-optical/Infrared (EO/IR) sensor and GA-ASI’s Lynx Synthetic Aperture Radar (SAR). Instead of streaming raw video or radar data back to a ground control station for analysis, the pod processes this information at the edge. This architecture allows the aircraft to identify objects autonomously, shortening the observe, orient, decide, and act cycle. By performing inference on the aircraft, the system reduces the communication bandwidth requirements needed to share target information with other platforms, a capability described by GA-ASI as critical for operating in contested or denied environments where datalinks may be jammed or severed.

The system was successfully demonstrated in September 2020 at GA-ASI’s Flight Test and Training Center in Grand Forks, North Dakota. During these flights, the Agile Condor Pod was integrated with the support of SRC Inc. to validate its ability to handle high-powered computing at the edge. The demonstration confirmed that the pod could accurately and effectively process sensor data to achieve information superiority. GA-ASI President David R. Alexander noted that the capability serves as proof that artificial intelligence technologies can lead to increased autonomous mission capabilities, allowing warfighters to operate more effectively in complex electromagnetic environments.

The Agile Condor is primarily associated with the United States military, specifically within the framework of the Air Force Research Laboratory (AFRL) initiatives. The pod is intended to foster the transition of AI technologies into operational constructs that improve the ability of unmanned systems to function independently. While the system has been demonstrated on the MQ-9 platform, specific details regarding its deployment in active combat zones or its integration into other platforms beyond the initial demonstration remain limited in public records. The technology represents a shift toward greater automation in unmanned aerial systems, focusing on rapid decision-making and target nomination at extended ranges.