The Shellcatch Virtual Observer System (VOS) pilot is progressing steadily in Barbados.
To date, three cameras have been installed at strategic points aboard the vessels Natalie, Morning Glory 2, and Runaway 1, and are actively transmitting data. After several trials adjusting the camera angles, an optimal position was found that provides a sufficiently wide frame to capture fishing activities (eg setting and hauling of gear).
Currently, the AI system is being adapted to recognise target species such as tuna and also flying fish, utilising landing photos and video footage. The system is also adapting to detect fishing activity based on cues such as vessel speed, human presence, and gear usage. While full implementation of AI-powered detection is taking place, manual review and annotation of footage remains necessary to ensure accuracy of the AI when in operation.
To improve visibility of specific operations such as the use of the down-the-line cutter, the Barbados Fisheries Division of the Ministry of Environment and National Beautification, are exploring the installation of a second camera. The current camera is focused on the loading area and will not adequately capture the down-the-line cutter in operation.
The Barbados team has participated in training sessions and follow up meetings with Shellcatch on platform navigation, technical troubleshooting, and AI training techniques. Ongoing efforts also include expanding the AI’s capabilities by identifying and registering additional species for recognition.
The purchase of five electronic monitoring (EM) units from Shellcatch Inc by the “Strategies, technologies, and social solutions to manage bycatch in tropical Large Marine Ecosystem Fisheries (REBYC-III CLME+)” project supports the Barbados Fisheries Division of the Ministry of Environment and National Beautification, Green and Blue Economy’s Sustainable Fisheries Management and Development Regulations which was enacted in 2024.
