New opportunities to learn about and help our lakes can arise unexpectedly. Such is the case when Madison Gas and Electric‘s Jeff Jaeckels reached out to Clean Lakes Alliance about a research group looking for volunteers to study the lakes in a novel way. Studying the lakes through artificial intelligence will allow Clean Lakes Alliance and our partners to better monitor and research harmful algal blooms.
The problem
Human-caused lake eutrophication and associated harmful algal blooms (HABs) are a challenge that extends well beyond the boundaries of the Yahara River Watershed. From neighborhood lake associations to regional public health departments, lake advocates are seeking efficient and cost-effective ways to test for water conditions that put people at risk. Lake conditions can change quickly, even over the course of a single morning. So, how can we test and notify communities about potential threats, or better yet, when the lakes are safe?
Let’s let computers do the work
Innovative researchers from New York-based team, BloomOptix, are exploring a way to train an artificial intelligence (AI) mobile app to quickly differentiate between harmless green algae and toxin-producing cyanobacteria. With the help of modern computing, BloomOptix hopes to solve the monitoring challenge. It will do this by using microscopic images of water samples to assess toxin presence. While still in early phases, the BloomOptix scientists hope to eventually provide lake managers a rapid identification tool. The goal is to provide a risk profile for the lakes using only a few pictures. A digital picture that can be taken with feet in the water and sent by a cell phone could have results in mere minutes. Gone would be the days of transporting lake samples from lake to lab to wait a day or more for actionable results.
Help from across the pond
But how do you take a microscopic image of algae while standing on the lakeshore? To overcome that problem BloomOptix enlisted the help of London-based company Iolight. Iolight manufactures portable digital microscopes that can conveniently link to a phone or tablet (Figure 1). The microscopes are small enough to fit in a pants pocket. They can take high-resolution images – a powerful tool for lake scientists and enthusiasts alike. Iolight plans to deliver approximately 300 microscopes across North America in 2022. This long-distance collaboration demonstrates the importance of looking beyond watershed boundaries to seek solutions to water quality challenges.
Volunteers get the ball rolling
Training AI requires large amounts of data to get up and running. As Igor Mrdjen, project lead for BloomOptix, explained, “We need pictures of algae, lots of pictures…thousands of pictures.” That’s where Clean Lakes Alliance volunteers like Arlene and Jeff Koziol and Robert Bertera step in. In fact, volunteers at six lakes across Wisconsin, including Mendota and Waubesa, helped collect microscopic images to send to the BloomOptix team (Figure 2). Throughout the summer, our volunteers sent close to 1,500 digital images for evaluation. Through the help of volunteer scientists, BloomOptix was able to collect large amounts of data otherwise not feasible by just their small team.
Thank you to volunteers and BloomOptix as they remind us that progress and innovation materializes when we work together.
