Small AI Models Enable Autonomous Drone Targeting on the Battlefield
NATO-backed startup Scaleout is deploying decentralized, small AI models to enable autonomous target identification and attack capabilities for military drones. This development represents a major shift toward decentralized edge AI in military technology, allowing autonomous systems to operate reliably even when electronic warfare jams external communication links. Scaleout leverages lightweight visual detection models like Ultralytics YOLO paired with federated learning, processing targeting tasks entirely onboard while keeping sensitive sensor data local and secure.
## BACKGROUND
Edge AI involves executing machine learning models directly on embedded hardware onboard localized devices rather than relying on central cloud servers. Federated learning complements edge computing by enabling decentralized devices to collaboratively update a shared AI model without sharing sensitive raw data across network channels.