For a few years now I am toying with the idea to create a "sniffing" device based on a fish tank. Multiple outlets around the sides, the diluted analytic soluti

For a few years now I am toying with the idea to create a "sniffing" device based on a fish tank. Multiple outlets around the sides, the diluted analytic solution is delivered on one outlet, the other outlets only carry tank water. The fish are trained to expect food at the site with a certain "smell". An overhead camera would detect the movement and distinguish it from random behavior. While fish are certainly able to smell minute concentrations, I never quite found a viable use case for this idea. For disease diagnostics, the idea of cumulating the pathogens in the fish and their water isn't particularly appealing either.