video1.0<iframe src="https://www.loom.com/embed/6d8f12b8901c4041aa3d9e2fe1ca523d" frameborder="0" width="1154" height="865" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>8651154Loomhttps://www.loom.com8651154https://cdn.loom.com/sessions/thumbnails/6d8f12b8901c4041aa3d9e2fe1ca523d-52fce683dce6cbe8.gif250.599In this Loom, I explain how our fusion engine scans for AIS custody failures when vessels go silent, treating silence as a detection event instead of dropping the track. We preserve last known state, assess what happened across the blackout using geography, weather, sanctions, contacts, and behavior, then refine confidence as new signals arrive. The interface scores dark periods into critical and high categories and shows them as color coded arcs on a static map, plus an intelligence object with hours dark, miles traveled, and predicted behavior. We also use an AI analyst to turn operator language into an actionable picture. No action was explicitly requested.