RISE ICT stories: Gaining full control of our coastline

Man looking at computer screens.

The Swedish Coast Guard is now deploying a data analysis module developed over several years at RISE SICS to detect problematic anomalies in vessel traffic along our coasts.

At any given moment, about 8,000 ships are operating off our shores. Approximately every ten seconds, data from each of these vessels is updated in a data analysis system, developed by SICS. The module analyses the data, looking for anomalies of several different types, based on either user-defined rules or historical data.

Currently, three types of anomalies are identified; movement patterns, meetings at sea and risk of grounding. When unusual movements by a ship are identified, e.g. turning unexpectedly or stopping too frequently, the alarm will go off. A drifting ship poses a great environmental risk and can be costly to rescue in a crisis situation. Movement patterns that are not typical to the boat type can reveal that the ship is doing something illegal. A "cruise ship" that behaves like a fishing boat may well be just that, attempting to evade fishing quotas. When two boats meet at sea, it may be indicative of fraudulent activity - usually smuggling, but it can also mean that fishing boats are unloading catches to circumvent fishing quotas. To check every ten seconds is a challenging computational task, which could not have been solved without the advanced techniques of big data analytics. Since the grounding of oil tankers is one of the most serious threats to our oceans, the possibility of real-time monitoring is a highly valued functionality. A vessel that moves out of a shipping lane and approaches waters shallower than the depth of the keel will receive an alarm about three minutes before the grounding will happen, which will give it time to maneuver away. Anders Holst, project manager at RISE SICS, believes that the long-term, close cooperation with the Coast Guard has given SICS a domain knowledge that makes the solutions adaptable to the real world and useful in practice. Furthermore, it will save taxpayers’ money.

Peter Ryman, Maritime analyst at the Swedish Coast Guard agrees: “One single rescue of a larger vessel, and the environmental consequences that are at risk, would outweight the cost of this project several times over”.

Read more

Big Data Analytics, Machine Learning and Optimization at RISE SICS

More RISE ICT stories