AZTI revolutionises the management and sustainability of pelagic fisheries – At the heart of technological innovation applied to commercial fishing, AZTI stands out as a pioneering leader. The research centre, renowned for its ability to integrate new technologies in the fishing industry, has recently developed an artificial intelligence (AI)-based application that revolutionises the management and sustainability of pelagic fisheries. This new tool is the result of advanced studies and applies AI technology to automatically classify the main pelagic species present in the Bay of Biscay, such as anchovies, sardines and Atlantic mackerel, by exploiting the behaviour of their shoals.
A step forward in fish stock management
The ability to classify fish schools detected by acoustic echo sounders represents a significant advance in the management of pelagic species, especially in an ecosystem as diverse as the Bay of Biscay. Thanks to an automatic classification model, it is possible to study species-specific behavioural changes in the presence of others, providing a more complete view of ecological dynamics. This approach fits within multidisciplinary investigations such as JUVENA, which are not limited to the management of fish species but aim to understand the functioning of the entire pelagic ecosystem, from plankton to large predators such as birds and cetaceans.
Artificial intelligence for species classification
The model developed by AZTI is based on poorly labelled data, using both fully and partially labelled schools of fish. The results, expressed in probabilities, make it possible to assess the reliability of the model in classifying each school of fish. This technological innovation is particularly promising: the use of acoustic data from sonar and echosounders installed on fishing boats improves the management of fish stocks. In addition to reducing data processing time during scientific surveys, the model increases accuracy in analysing the distribution and annual abundance of pelagic species.
Sustainability and by-catch reduction
The automatic classification of fish shoals has significant potential to improve the efficiency and sustainability of commercial fishing. A major challenge for the industry is the reduction of by-catch, i.e. the accidental capture of unwanted species. The technology developed by AZTI not only significantly reduces by-catch, but also contributes to making fishing more selective and, consequently, more sustainable.
Results published in the ICES Journal of Marine Science show an accuracy of 63.5 per cent when classifying tagged schools and around 80 per cent when considering tagged and untagged schools at the fishery level. These numbers demonstrate the success of a system that promises to transform the way pelagic fisheries are managed, providing new opportunities for the study of shoal behaviour and the development of technologies to improve the efficiency and sustainability of the industry.
AZTI Technology leadership and sustainability
AZTI confirms itself as a benchmark in the sustainable fishing landscape, thanks to its ability to develop tools that combine machine learning, big data and AI. An example of their commitment is their participation in the European project SMARTFISH, aimed at developing intelligent systems for the EU fishing sector, with the objective of reducing the environmental impact of fishing. In addition, AZTI leads expert groups in size and species discrimination for purse seine fisheries, emphasising its crucial role in promoting more efficient and sustainable practices.
With these initiatives, AZTI is pioneering the application of artificial intelligence and big data to fisheries, demonstrating how technological innovation can contribute significantly to the sustainability of marine resources.
AZTI revolutionises the management and sustainability of pelagic fisheries
