Over 90% of the world's trade is carried by sea and is, by far, the most cost-effective way to move goods and raw materials
around the world, but the current ship tracking system, the Automatic Identification System (AIS), relies on ships and
vessels being “co-operating vessels”.
Ships that are not voluntarily disclosing their position (so-called “non-cooperative vessels” or “dark vessels”) are more
often involved in criminal, illegal and illicit activities, and they only have to flip a switch to become "invisible".
Detection of dark vessels is possible with other technologies, e.g. SAR, but this technology is restricted because of a
narrow image area, required data processing to render images, and a complex processing to detect anomalies that cause
false alarms. With multispectral images it is possible to apply deep neural networks (machine learning) to automatically
detect, classify and identify ships, ship sizes and heading, and it is possible to do this much faster, for much larger areas
and to obtain more detailed information about the vessels found in these images.
Vake will build the system that combines the information sources, yielding a much better, global overview of all vessel traffic.