Introduction
R-Control develops generic methods and technologies that enable estimating the catch composition arriving on the fishing vessel, applicable to multiple fisheries. The catch composition from each fishing effort will automatically be analyzed regarding species composition and size composition of each species in terms of length and weight. The capture information will automatically be linked to the time and location of each fishing effort, gear used, and other parameters regarding the fishing process.
Our goal is to place Norway at the forefront of research and development of core technology for effective and sustainable resource control in commercial fishery.
The main hypothesis of this project is that the implementation of automatic catch registration systems in commercial fishery can lead to a more effective resource control, precise resource management at lower costs while simultaneously providing decision support for fishers and managers leading to increased compliance with regulations.
For automatic catch registration to be both widespread and accurate, there must be an efficient way of generating large data sets for use in deep learning algorithms. This project addresses the generic methods and technologies that enable generation of such large data sets using the principle of anatomically correct digital twins of fish, along with modern data set generation methods such as domain and dynamics randomization.
Automatic capture information and catch registration technology will lead to a full-documented fishery that will:
•Enable automatic check for compliance with regulations regarding species and size compositions of catchand bycatch.
•Enable certification to consumers on sustainability of the fishing process.
•Provide decision support for the fishers in between fishing efforts.
•Enable information for real-time decisions regarding potential area closures to maintain sustainability.
•Provide data for stock assessments directly from the commercial fishing activity.