Biological Sciences: Fully Funded PhD Scholarship: Novel machine-vision and deep-learning applications to assess the welfare of cleaner fish in the salmon industry
Funding providers: VisiFish Ltd. and Swansea University's Faculty of Science and Engineering
Subject areas: Behavioural Ecology, Aquaculture, Machine Learning, Fish Conservation
Project start date:
- 1 July 2022 (Enrolment open from mid-June)
Project description:
Fish welfare is central for the sustainable growth of aquaculture but measurements are difficult to standardise due to the diversity of farmed species (more than 580). The ballan wrasse (Labrus bergylta) and the lumpfish (Cyclopterus lumpus) are currently used as cleanerfish (a biological alternative for parasite control) in the UK salmon industry. The three species do not coexist in the wild but are reared together in salmon cages, raising concerns about their health and welfare. This PhD project will combine fish biology (Swansea), machine vision and deep-learning algorithms (FAPTIC) to develop and optimise tools to assess and improve the welfare of cleanerfish in salmon cages. This will make the long-term use of cleanerfish in the salmon sector more sustaibnable and allow Wales (the UK’s largest producer of cleanerfish) to further grow their industry. The project will focus on the following 3 areas:
- Baseline testing of fish interactions at the facilities in CSAR
- Tool development to assess, maintain and optimise cleanerfish welfare through machine learning
- Validation in industry
The Centre for Sustainable Aquatic Research (CSAR) team is a UK leader on cleanerfish research and has helped develop the Welsh industry. They have also pioneered the development of welfare assessment tools for lumpfish. This project will expand their experience on novel tools for welfare assessment in partnership with VisiFis, a pioneer company using telematics and AI tools to monitor mortality, individual and population behaviour and growth rate. In collaboration with CSAR at Swansea University and VisiFish the student will develop tools for early detection of anomalies in fish welfare and behaviour in fish farms. This early notification is crucial to preserve fish welfare and avoid commercial losses.