Lockheed Martin announces revolution in SAR

Lockheed Martin has revolutionised maritime surveillance with a recent development that enables a sensor using synthetic aperture radar (SAR) to autonomously detect targets, the company said in a July 16 press release. 

Using an AI-powered SAR with embedded sensor control autonomy during a flight test on the West Coast of the US, the technology has the “potential to detect and track maritime targets with unprecedented accuracy,” the company claims. During the demonstration, the SAR Automatic Target Recognition (ATR) capability, which is enabled by Machine Learning Operations (MLOps) tools for rapid retraining, successfully classified targets of interest in near real-time, the press release detailed. 

It added that SAR ATR was also coupled with autonomous sensor control, which re-tasked the radar in real time based on target detections. The press release specified that this capability was deployed on low Size, Weight, and Power (SWaP) hardware in the field, utilising rapid edge processing without the need for large cloud computing or ground stations.

Testing of the AI-powered SAR ATR will continue throughout the year and data collected from these flight tests will be used to develop and mature multiple Lockheed Martin autonomous systems, including collaborative combat aircraft and broader family of systems applications, the press release said. 

As previously noted in other Calibre Defence articles, SAR is like a highly advanced camera that uses radio waves instead of light to create detailed images of the ground or sea. Unlike a regular camera that needs light and clear weather, SAR can “see” through clouds, darkness, and even some obstructions because radio waves can penetrate them. It works by having an antenna, often on an aircraft or satellite, send out many short radio pulses as it moves. By carefully collecting and combining the echoes from these pulses, and understanding how the aircraft moved, powerful computers can create a much sharper image than a regular radar could.

The development was described as “a major leap for harnessing AI to help enhance situational awareness and decision-making capabilities, with unparalleled threat identification across extended ranges and all-weather conditions,” by John Clark, senior vice president of technology and strategic innovation at Lockheed Martin. 

Calibre comment

It is obviously difficult to assess what Lockheed Martin has achieved on the merits of its own press release. However, we can say with confidence that this is not the first time that AI has been applied to the field of SAR. In November 2023 Agenium Space, a French company specialising in space-based AI, reported that it had successfully deployed its deep neural networks to Loft Orbital’s YAM-3 satellite, where they conducted object detection on synthetic aperture radar (SAR) imagery. Loft Orbital has partnered with Helsing to add AI to its satellites in support of SAR target detection. ICEYE has also partnered with Safran.AI to add AI into its product offering, which includes dozens of SAR satellites. This indicates that applying AI to SAR imagery is not particularly novel or trailblazing. It certainly is effective and delivers better results in shorter time than a human analyst, but it is unclear what, exactly, is meant by the Lockheed press release insofar as revolutionising maritime surveillance is concerned. There are certainly plenty of drones running AI on the edge for optical sensors; so perhaps it is the application to SAR that is revolutionary here? 

By Sam Cranny-Evans, published on July 18, 2025.