Applying deep learning to predict and mitigate adverse environmental conditions in fish aquaculture pens

Fish farms (Aquaculture) produce a large and growing share of seafood in Canada, 29% by value and 19% by production. Worldwide, aquaculture now surpasses wild fish catch. We need aquaculture to be safer for the environment, low cost and yield more fish to feed a growing population in the face of climate change. A key driver of fish health, fish growth, and fish feeding activity is a healthy environment, often measured by dissolved oxygen in water. In this project we will use large datasets from multiple aquaculture pens to train deep learning models to predict future dissolved oxygen levels and other metrics to help predict the impact of mitigation efforts. This work will help provide better decision making that helps reduce the energy use, cost, and environmental impact of aquaculture.

Intern: 
Vegu Shree Rama Kamal Kumar
Faculty Supervisor: 
Christopher Whidden
Province: 
Nova Scotia
Sector: 
Partner University: 
Discipline: 
Program: