Using Machine Learning for Ocean Data Analysis through integration of satellite images and ship movement data

This project aims to combine two different sources on merchant ship movement, AIS and satellite images. AIS data is emitted by every merchant ship above a certain size, which allows the ship to be tracked from the coastline. The goal of this project is to use both these sources of data to train an automated system to detect suspicious ship movement, that could indicate illegal behavior (e.g. illegal transfer of cargo), and continue tracking ships via the imagery even when the AIS transmitter is (purposefully) turned off and the ship is no longer emitting its position. This is a novel approach, as typically in the past, although several different ways exist to track merchant ship movement, the data from these systems is rarely used in combination to achieve a more in-depth understanding of shipping traffic, and tends to be analyzed in isolation. The expected outcome of the project is an automated system that can identify specific ships from satellite images and determine discrepancies between the two sources.

Faculty Supervisor:

Stan Matwin

Student:

Partner:

University of the Aegean

Discipline:

Computer science

Sector:

Ocean Tech; Information and Communications Technology; Transportation (excluding aerospace)

University:

Dalhousie University

Program: