Safe Pro Wins New U.S. AI Subcontract, Launches Ground Intelligence Suite

Safe Pro Group said it won a new U.S. government subcontract and launched SPAI Ground Truth, a five-part AI suite built to fuse air and ground sensor data for battlefield threat detection. The company says the system is based on more than 3.1 million drone images from Ukraine and is designed to help soldiers detect mines, munitions and ambush drones in real time.

Why it matters: - Safe Pro is targeting a battlefield problem that is getting harder to solve with drones alone: fast-changing threats on the ground layer. - The company says SPAI Ground Truth is built to improve situational awareness for soldiers, which could affect route planning, maneuver operations and threat detection. - The new U.S. government subcontract gives Safe Pro another foothold in defense work tied to AI autonomy and threat detection.

What happened: - Safe Pro Group Inc. said it won a new U.S. government subcontract and launched SPAI Ground Truth, an integrated software suite for battlefield awareness. - The company also said it received an initial order from a new U.S. government prime contractor customer for its patented AI software. - The announcement was made Oct. 8, 2026.

The details: - SPAI Ground Truth combines air and ground AI algorithms to share real-time threat information from multiple sources, including drones and unmanned ground vehicles. - The suite includes five components: GT Cloud, GT Air, GT Road, GT Field and GT NODE. - GT Cloud is AWS-hosted and handles scalable processing, AI threat analysis and 2D/3D mapping. - GT Air runs on unmanned aerial systems and performs real-time in-flight AI threat detection. - GT Road runs on ground vehicles and UGVs and detects anti-vehicle mines while in motion. - GT Field runs on laptops and tablets and supports real-time and post-flight AI detection in drone imagery and video. - GT NODE is a portable edge-compute server that enables fully offline post-flight AI detection and 2D/3D interactive threat mapping on the edge. - Safe Pro says the suite can identify more than 150 types of explosive threats and objects of interest, including landmines, cluster munitions, unexploded ordnance and ambush drones. - The system produces 3D models, orthomosaics, digital surface models, and vegetation-height and terrain-slope layers with detected threats mapped onto them. - Outputs are designed to be shared over existing tactical networks, including the ATAK/TAK ecosystem. - The company says SPAI Ground Truth is based on more than four years of operational use and data collection, including more than 3.1 million real-world drone images from Ukraine. - Safe Pro says its AI processed more than 3.1 million drone images, confirmed 58,526 threat detections and mapped more than 39,777 acres in Ukraine. - The company said additional terms of the subcontract were not provided.

Between the lines: - The product launch shows Safe Pro expanding from drone-image analysis into a broader system that links air, ground and edge computing. - The Ukraine dataset is central to the company’s pitch, because it provides a large real-world training base for threat detection models. - The focus on offline and edge capabilities suggests Safe Pro is aiming at environments where connectivity is limited or contested.

What's next: - Safe Pro said it expects SPAI Ground Truth to support defense, humanitarian and homeland security applications. - The company also signaled potential follow-on, production or other awards tied to the subcontract, though no timing was given. - More information is available in the company’s landmine and UXO detections page and on Safe Pro Group’s website.