Scaleout Systems, a company founded in 2018 by Uppsala University researchers, has shifted its focus to defense technology following the outbreak of the war in Ukraine. The startup is currently supported by NATO and participates in the Defence Innovator Accelerator for the North Atlantic (DIANA) program.
The company develops lightweight machine learning models designed to run directly on edge hardware, such as drones and field workstations. These models enable real-time target detection and identification without requiring a constant connection to centralized servers, mitigating risks from electronic warfare and signal interference.
Scaleout Systems is involved in the ALMA project led by BAE Systems Bofors, which aims to create low-cost autonomous loitering munitions. A prototype demonstrated in January 2026 showed that drones could autonomously detect, identify, and strike high-priority targets like armored vehicles using onboard computing.
The system utilizes a federated learning architecture, allowing local devices to continue AI inference and model training even when communication with command nodes is lost. Once connectivity is restored, updated models are synchronized back to the central network, enabling continuous iterative improvements.
