Two Smartfish participants Helen Holah (Marine Scotland Science) and Kristian Plet-Hansen (DTU Aqua) presented at the Seafood Conference Iceland (Reykjavik, Iceland) on the 07th November in a session titled ‘Remote electronic monitoring (REM) in fisheries’. The session was sponsored by the Nordic council of ministers working group for fisheries (AG-fisk) with the purpose of starting a discussion in Nordic countries about increased surveillance in fisheries. In a similar fashion to the implementation of the Landing Obligation within the European Union the main focus of discussion on fisheries control in the Nordic countries in recent years has been management of bycatch. In Iceland for example a draft legal bill on increased digital surveillance was presented for consultation online in July 2018, which has catalysed a lively debate on the pros and cons of REM. Currently Icelandic and Norwegian fisheries fall into the 45% of global landings from fisheries with few or no data on discards as described by the FAO.
Conference attendees heard about experiences of the application of REM within the fisheries of Chile, Canada, the US, Denmark and Scotland. Some of the topics covered included; What are fully documented fisheries? (FDF), what is Electronic Monitoring? (EM), system set ups, video auditing, cost comparisons, and automated image recognition software and technologies. The application of EM was discussed in order to increase Monitoring Compliance and Surveillance (MCS) of regulations regarding discard bans, Catch Quota Management (CQM) trials and Landing Obligations (LO).
The presentation ‘The evolution of a commercial fishery: Experience of fishermen working under REM’ by Wes Erikson, a commercial fishermen and director of the Halibut Advisory board in British Columbia, prompted interesting discussions surrounding incentives vs ultimatums in implementing EM. Initial installations of EM for time limited ‘research programmes’ prior to introducing sanctions were believed to have helped build trust in both Chilean and Canadian EM programmes. The use of independent non-governmental professional facilitators or contractors to grease the wheels of industry/science/manager interactions or manage and analyse EM data was also highlighted as an aid to success. The conclusion that fully monitored fisheries eliminate questions of trust and allow industry to build a transparent relationship with management, science and the public and the metaphor that it can be considered paying for an insurance with which a fishery is defensible demonstrated positive transformations in thinking.
Helen Holah in her presentation titled ‘Electronic monitoring of Scottish fisheries: a strategic approach to developing an evidence base’ described the training data set built
by Marine Scotland Science for utilisation by the University of East Anglia School of Computer Science in the development of automated image analysis of fish processing conveyor belts on Scottish demersal fishing trawlers. Going on to explain how this project has become the CatchMonitor deliverable of work package 4 (Development of automatic catch analysis systems) of the Smartifsh project which through on-board catch monitoring using CCTV systems will verify the number, size and species of caught and discarded fish using deep learning algorithms. The output of this will be displayed on a console as well as storing data outputs and images to enable subsequent analysis by fisheries agencies onshore. The industry stakeholders were particularly interested in the current readiness state of AI technologies as both the time taken and cost of video analysts to review EM footage by eye is one of the largest perceived cons.