SCOUT

The SCOUT project develops an integrated UAV swarm platform that combines Integrated Sensing and Communication (ISAC), computer vision, and edge/cloud intelligence to provide robust situational awareness in remote environments. The system addresses three civil use cases of high societal relevance: search and rescue operations, precision livestock farming, and wildlife monitoring.

A core innovation of SCOUT lies in its multi-modal target detection system, which fuses complementary sensing modalities. ISAC-based methods analyze variations in wireless signals to detect motion and vital signs, while a vision-based model (VisionAware) employs deep learning on RGB and thermal imagery to identify humans and animals under diverse environmental conditions. Sensor fusion algorithms integrate both modalities, ensuring reliable performance in low-visibility or cluttered scenarios.

To support collaborative detection, SCOUT introduces a dual-network UAV communication backbone. A long-range, low-throughput LoRa link maintains robust command-and-control connectivity, while a short-range, high-throughput network (5G NR sidelink or DSRC) enables real-time data exchange and cooperative sensing among UAVs. The resulting architecture provides infrastructure-independent operation, critical for deployment in rural or disaster-stricken areas. The platform will be validated through laboratory integration and field trials, with performance benchmarks including detection accuracy, latency, false positive rate, and energy efficiency. By advancing trustworthy AI, resilient wireless networking, and modular UAV-based sensing, SCOUT contributes to the development of next-generation autonomous systems for civil applications.

Project duration: 01.03.2026 - 31.05.2029

Consortium:

• Luxembourg Institute of Science and Technology
• University of Luxembourg
• BlueSpace Technology
• SDR Engineering

Funding: Luxembourg National Research Fund (FNR), INTER/EUROSTARS25/19825335/SCOUT.