Skip to main content
Citizen science has long been dismissed as unreliable. But as leading Canadian charities like Swim Drink Fish utilize machine learning technology, that's changing.

You may have heard of Swim Drink Fish. Operating since 2001, it’s a Canadian charity with a mission that is hard not to get behind: “working for a day when every person can safely touch the water, when the water is pure enough to drink, and when the water is clean and wild enough you could toss a line in and pull out a fish.”

Using the latest technology, the charity works on data collection and public awareness to restore and inspire care of Canada’s water.

When the charity launched, buying the latest technology — a good camera or a GPS device — was expensive. Fundraising could take months, and often high-priced gear couldn’t be operated without some expertise.

“Now there’s a huge transformation where most of us have a great camera and GPS tracker embedded in our phones,” says Krystyn Tully, Co-Founder and Vice President of Swim Drink Fish.

An App for Cleaner Waters

It didn’t take long for Swim Drink Fish to make the most of people’s connection to their smartphones. In 2011, the charity launched its Swim Guide app, which lets users know which beaches are or aren’t suitable for swimming.

Within the first six weeks of launch, there were approximately 20,000 downloads. Tully says, “It was then that we realized we were onto something.”

During its first summer, the app expanded from Toronto, to the entire Great Lakes region, and then to British Columbia. Now, in 2020, the app is in eleven countries, available in 3 languages, and has had 4 million users.

These days, another exciting technological push is happening at Swim Drink Fish. Tully explains that, with advances in smartphone technology, it’s now the case that “ordinary people have the ability to use their phones to do citizen science.”

Citizen science is where research is gathered by people who aren’t professional researchers. To take a famous example, when thousands of people take part in the Christmas Bird Count each year by recording the number and types of birds in their area over a single day — they’re being part of a citizen science project.

But say a person is by Lake Ontario’s shoreline on a Monday morning, and they notice the water has a surprising amount of trash in it. That may be a good observation, but how can that information be used?

Through support from a RBC Tech for Nature donation, Swim Drink Fish is working on the machine learning tools to quickly and accurately interpret such observations into scientific data.

If the observer takes a photo of what they’re seeing and uploads it to the Swim Drink Fish app in the future, a machine learning image recognition tool could be used to accurately and efficiently deduce if the water near the shore is unusually full of trash. It could even identify the different elements of litter, from chip bags to pop bottle lids. Say the image is unusually full of plastic bottles, staff at Swim Drink Fish can come up with a plan of action. Noticing, for example, there was a festival over the weekend, they can reach out to festival organizers to come up with plastic alternatives at next year’s event.

If thousands of citizen scientists across Canada upload photos of their local waterways to the app for analysis, the impact could be huge.

“RBC has been supportive of the process right from day one,” Tully says. “And because of their help — supporting the us through funding and employee volunteering, where our leadership team collaborates with their employees to solve business planning and fundraising issues — Swim Drink Fish looks set to use cutting-edge technology to help protect water for a long time yet.”

That’s good news for everyone. And it really turns the notion that nature and technology are incompatible on its head. “Socially, there is a growing disconnect from nature,” says Tully. “Our whole philosophy is, can we use technology to be part of the solution?”