Researchers developing AI to estimate amount of waste entering Sask. landfills

Researchers developing AI to estimate amount of waste entering Sask. landfills

A Regina-based technology firm is working with University of Regina researchers to develop artificial intelligence that will estimate the weight of solid waste entering Saskatchewan landfills.

The idea is to use machine learning to determine the types of waste and how much of it is being deposited in those landfills in the province not equipped with expensive weigh-in-motion technology.

The AI will find correlations between the images captured on cameras at those landfills and the actual weight and composition of waste entering those landfills, then use that information to estimate the content and weight of waste received at remote and rural landfills that will only have the cameras.

Prairie Robotics will develop the technology, which will capture data in real time using cameras sturdy enough to handle Saskatchewan's weather and automatically generate up-to-date waste reports across the province.

One of the researchers for Prairie Robotics, Joshua Friedrick, said this technology will help Saskatchewan Environment manage waste.

An image, created by Prairie Robotics for a proposal, showing the potential of what artificial intelligence can classify in terms of waste vehicles and waste types. (Prairie Robotics)

"Without up-to-date accurate data, your decisions are somewhat volatile or they're susceptible to the human bias," he said.

"So with using machine learning to quickly and accurately capture the data that they need, it enables them to make better decisions for all of Saskatchewan, including the rural sites, which is a bit of a challenge right now using their current technologies."

Colleague Sam Dietrich said this approach has been used elsewhere and will help governments make informed decisions.

"Let's say a lot of concrete is being received in southwest Saskatchewan," he said. "Well, if a lot of concrete is being received there and they're looking for a recyclist, with that data they could say, 'You know where a great place for a concrete recycling plant would be? It would be in southwest Saskatchewan where this waste won't have to be moved great areas.' "

Their work has attracted the attention of the provincial government.

The researchers have been selected as the winners of Saskatchewan's latest Innovation Challenge.

Each of the winners will receive funding of up to $10,000 and a 16-week residency to develop their solution in collaboration with government.

[Source: CBC.ca]

"The data that this type of technology can provide will be extremely useful as we determine the effectiveness of our waste diversion programs in the province," Environment Minister Dustin Duncan said in a release.

The researchers have been given until March 31, 2020 to complete their residencies.