MONITORED AI

MONITORED AI is a cloud-based maritime surveillance platform from OPT/NET that detects abnormal vessel behaviour by fusing SAR satellite imagery with AIS ship-tracking data. The detection pipeline creates composite geo-images from SAR scenes, suppresses background clutter, and applies two vessel-detection models - one for large and medium vessels and a second tuned to smaller, non-metallic targets. Detected SAR contacts are matched against AIS transmissions using azimuthal correction and cross-correlation; unmatched vessels (dark vessels without active AIS transponders) are flagged as anomalies and clustered for operator review. ^[1] SAR input sources demonstrated through the ESA ECDPIF project include Sentinel-1, TerraSAR-X, PAZ, and ICEYE. Results are delivered through a graphical user interface with chatbot assistance and are exportable in GPX, KML, and GeoJSON formats, with per-vessel records including vessel name, MMSI, and photograph where available. ^[2] An MVP was developed for Dutch maritime safety authorities in partnership with the Dutch police. MONITORED AI is also a component of the MARTINA consortium, a European Defence Fund programme led by NLR for satellite imagery analysis applied to security and defence. ^[3] As of 2026, OPT/NET website does not list commercial licensing terms for MONITORED AI outside government and defence procurement.