AI-powered military systems often rely on cloud networks, but in combat zones, latency, jamming, or disrupted connections can delay critical decisions.
Lantronix and Safe Pro Group have partnered to address this issue, embedding AI directly onto unmanned systems to enable real-time threat detection even offline.
The collaboration combines Lantronix’s compact onboard computers with Safe Pro’s AI-driven computer vision software, allowing drones to analyze sensor data on the fly and flag potential hazards immediately.
Onboard Processing and Integration
At the core is Safe Pro’s Object Threat Detection (SPOTD) AI model, which scans aerial imagery to identify landmines and other small hidden threats.
Processing runs locally on Lantronix’s Open-Q 5165RB module, powered by a Qualcomm edge-AI chip.
Detection results can be routed directly into existing military networks, including the Android Tactical Assault Kit (ATAK), and integrated with unmanned platforms already in service, such as drones supplied by Red Cat Holdings.
Dan Erdberg, chairman of Safe Pro Group, said the partnership with Lantronix is “empowering real-time intelligence at the edge for a wide array of defense and commercial applications, creating immediate synergies for Safe Pro Group.”
