Underwater Monitoring Using Sonars and Deep Learning
High volumes of jellyfish regularly clog the pipes conducting cool water to the turbines of power stations which forces the latter to shutdown. It stops plants from generating power and might take up to several days to clear the filters.
The aim of the project is to use imaging sonars to gather data in the underwater marine environment and develop Deep Learning algorithms that will be capable of processing and analysing data on small embedded platforms without consuming too much power. That way, it will be possible to warn the power stations when jellyfish are present around in an unmanned manner.