EDF-Store, The European database for the development of Military Optronics AI

Optronics sensors, coupled with advanced image recognition technologies, are vital for modern defence platforms facing complex, rapidly evolving threats. The STORE project, a European initiative uniting industry and research organizations from nine countries, addresses critical challenges limiting AI deployment in defence image recognition, notably the scarcity and classification of representative datasets. This paper details STORE’s objectives, approach, and main
innovations: a secure, distributed and federated database for data and model sharing, a unified AI framework for collaborative development and benchmarking of AI functions, and defence strategies for private, federated and secured
learning of AI approaches. STORE aims to accelerate the readiness, robustness and interoperability of AI-powered image recognition systems, providing enduring operational advantage to European defence stakeholders. 1. INTRODUCTION
The emergence of new high-intensity threats—such as hypersonic missiles, drones and drone swarms, and advanced loitering munitions—presents formidable operational challenges to modern armed forces. Efficient engagement, force protection, and
tactical decision-making increasingly rely on timely, accurate, and robust interpretation of large volumes of multispectral imagery captured by networked optronic sensors. Artificial intelligence (AI) and, more specifically, machine learning (ML) methods now underpin a growing share of image-processing pipelines in defence, promising substantial advances in target detection, recognition, and situational awareness. Yet, across Europe, the operational exploitation of these technologies remains hindered by the limited availability of representative, annotated datasets
and by a lack of common standards for secure and efficient AI, data, and model sharing across organizational and national boundaries.