BUOYS THAT LISTEN
USING TARINGA TO DETECT ILLEGAL FISHING
Illegal fishing is an urgent issue for the health of the world’s ocean, including here in New Zealand. Currently, detection is largely reliant on coastal patrolling, which is both labour intensive and costly. There is a significant need to develop new ways to detect illegal activity, in a simple to deploy and affordable manner. This innovative project brings together existing work by leading New Zealand underwater acoustic technology company, Ocean Instruments, with marine and machine learning scientists from the University of Auckland and Ocean Acoustics. It aims to develop a system that uses sound to detect illegal fishing and send real time notifications to apprehend offenders.
Underwater sound travels long distances, so Ocean Instruments and Associate Professor Craig Radford from the University of Auckland collaborated to develop an acoustic buoy that transmits underwater sound back to shore. The system is called Taringa, Māori for ear. They’ve already collected thousands of hours of acoustic data and have developed a high precision machine learning algorithm to detect the presence of vessels. Now, the aim is to classify vessels by sound in partnership with Dr Yun Sing Koh Associate Professor of Computer Science at the University of Auckland and Dr Matthew Pine from Ocean Acoustics. Live Ocean is supporting specialist Amol Malla for this boat detection and classification work, using AI.
The marine reserve at Goat Island will be the initial testing ground. It’s a no take marine reserve, managed by DOC, and the aim is to detect boats fishing within the reserve at night, when most illegal activity occurs. The system will log vessel activity each night and patterns can be examined for signs of illegal fishing. The ultimate aim is to be able to send real time notifications, alerting authorities when regular activity is detected.