Autonomous strike drone uses Nvidia Jetson Orin Nano to pick and bomb targets (www.tomshardware.com)

A recent demonstration by Swedish AI startup Scaleout Systems showcased an autonomous strike drone that utilizes Nvidia's Jetson Orin Nano to autonomously identify and engage targets. This low-cost loitering munition conducted a successful mission during BAE Systems’ Winter Demo 2026, where it detected, geolocated, and attacked an armored engineering vehicle with minimal human intervention. The drone operated entirely onboard, processing information without needing communication links, which demonstrates resilience against electronic warfare. The mission was completed in under 320 seconds, with around 200 seconds spent on reconnaissance, highlighting the efficiency of the system. Significantly, Scaleout employs a unique "Tactical Computer Vision Network" which utilizes federated learning for local AI training, allowing the drone to adapt rapidly to various environments. With its capabilities demonstrated even in extreme conditions, such as sub-zero temperatures, the technology indicates a future potential for autonomous systems in military operations. However, concerns surrounding lethal autonomous weapons are prompting a review by a UN expert group, emphasizing the need for human control in compliance with warfare laws. As Scaleout aims to transition from demonstration to potential NATO procurement, the balance between innovation and ethical considerations in autonomous warfare remains a critical focus for the AI/ML community.