Intelligent transportation system for arrival time and vessel speed predictability

XST is a private company with the exclusive mandate to manage and develop the Canadian Coast Guard’s Vessel Traffic Management Information System, which is the federal government’s official vessel traffic database. XST also has unique and exclusive international sales rights on this system.

Context of the project

The project aims to develop a web platform that will be accessible free of charge and will share an improved version of estimated arrival times (ETAs) and vessel speeds. This data will be generated by predictive artificial intelligence models based on machine learning. The models will be developed from INNAV data history that contains over twenty years of navigation data. Currently, the ETA predictions of the vessels found on INNAV or used by platforms available on the web (Marine Traffic, VesselFinder, Fleetmon, etc.) are based on average and constant speeds attributable to the various navigation sections through which the vessels transit. With that method, the ETA prediction several hours before the arrival of vessels is unreliable, which results in an efficiency loss for port authorities. Actuellement, la prédiction de l’ETA des navires que l’on retrouve sur INNAV ou qui est utilisée par les plateformes disponibles sur le Web (Marine traffic, Vessel finder, Fleetmon, etc.) est réalisée sur la base de vitesses moyennes et constantes attribuables aux différents tronçons de navigation par lesquels les navires transitent. Avec cette méthode, la prédiction de l’ETA plusieurs heures avant l’arrivée des navires manque de fiabilité, ce qui constitue une perte d’efficience pour les autorités portuaires.

Positive fallout

The project will make it possible to calculate a much more accurate ETA and predictive speed of vessels seven days before their arrival at a port in Québec. The geographic area covered by the project includes the Great Lakes, as well as the limit of Canadian waters in the Atlantic. All vessels bound for or operating on the St. Lawrence River or Saguenay River are targeted by the project. 

The full development of the project, whose final delivery is a public web platform that includes artificial intelligence-driven predictions of ETAs and ship speed, is scheduled for October 2025. This is a project that will therefore be implemented over approximately three years.