Coast Guard Headquarters Develops AI-Based Automatic Detection System for Drunk Sailing Vessels

Coast Guard Cracks Down on Drunk Boating
Photo by Mokpo Coast Guard

[Asia Economy Reporter Park Hyesook] The Korea Coast Guard announced on the 9th that it has developed a system that automatically detects vessels suspected of drunk driving at sea based on artificial intelligence (AI).

This system operates by automatically detecting vessels with navigation tracks similar to those of previously caught drunk-driving vessels and notifying vessel traffic controllers.

To develop the automatic detection alarm system, the Coast Guard used AI to analyze 221 navigation track characteristics of drunk vessels, such as zigzag navigation, from 551 cases of drunk driving enforcement and accidents since 2016.

Previously, the Coast Guard viewed it as virtually impossible to monitor vessels freely navigating the vast sea with limited patrol resources, so it promoted the development of an alarm system using VTS resources that monitor operating vessels 24 hours a day.

Then, in July last year, it established the venture-type organization 'Seonbak Gyotong Gwanje Techildan' (Vessel Traffic Control Technology Team) and completed the development of the alarm system after a year of project promotion.

In particular, from data collection and analysis for technology development to AI development, all were performed with in-house technology, resulting in a significant budget reduction effect.

The Coast Guard plans to first install and operate this system at Mokpo Port VTS to verify the effect of preventing drunk driving, then improve its accuracy and expand its operation nationwide.

A Coast Guard official said, "We have now taken the first step in developing marine accident prevention technology using AI," adding, "We will continue to develop technology to create safe sea routes."

Meanwhile, drunk driving accidents have occurred between 11 and 22 cases annually from 2016 to last year and are on the rise.