GoKawiilScaleout Systems demonstrates NATO-linked loitering munition that autonomously selects and strikes targets
Swedish startup Scaleout Systems showed off an autonomous loitering munition, built on a BAE Systems Bofors airframe, that used onboard AI running on an Nvidia Jetson Orin Nano to detect, rank and strike an armored vehicle without human guidance mid-mission. The demo, part of NATO's DIANA-backed ALMA and FEDAIR programs, completed the full flight in under 320 seconds using no external communications, relying instead on locally trained computer-vision models shared via a 'Tactical Computer Vision Network.' A human pilot remained available as a failsafe but was not required once the mission began.
GoKawiil's interpretation of the reporting above, not reported fact.
This shows lightweight, non-frontier AI models can already handle target identification and engagement decisions onboard cheap drones, lowering the technical bar for autonomous weapons deployment. Requiring zero external communications also makes such systems resistant to jamming and electronic warfare, a major concern in current conflicts, while raising fresh questions about human oversight in lethal targeting decisions.
- Scaleout Systems' drone used a small onboard AI model on Nvidia Jetson Orin Nano hardware to autonomously detect and strike a target.
- The system needed no external communications, completing its mission in under 320 seconds and resisting electronic warfare interference.
- The project is backed by NATO's DIANA accelerator and BAE Systems Bofors under the ALMA and FEDAIR programs.
Source: tomshardware.com — Shane Downing, 2026-09-20
Published there as: “Autonomous NATO strike drone uses Nvidia Jetson Orin Nano to independently pick and bomb targets — Swedish startup's attack drones run small AI model, require no human input and zero external comms”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.