Canada leads the developed world in the production of garbage. Every citizen discards 720 kilos of waste annually—twice as much as the Japanese generate. Canada even tops the U.S.—the poster child for overconsumption—by seven percent.
Even worse than the sheer amount of waste produced, however, is how the country deals with it. Like many western societies, Canada ships much of its problematic garbage to poorer nations to bury. Although China had been the destination of choice, its recent ban on accepting foreign garbage has countries scrambling to find new places to send trash—with India and Pakistan at the top of the list.
For Pakistani-born Hassan Murad and Indian national Vivek Vyas, both grads of SFU, solving the waste crisis is personal.
After cofounding the startup Intuitive AI, the pair used their knowledge of artificial intelligence from internships at heavyweight companies like Tesla and International Submarine Engineering to come up with an innovative solution.
“When you walk up to a bin, you see four or five options to recycle,” Murad tells the Georgia Straight by phone during a break between investor meetings in Toronto. “You look at the labels, and they’re really confusing. One might show a coffee cup, but then the sticker on the product might say it’s compostable. And then you get more confused if there’s still liquid or something else inside. That’s why we created Oscar, an AI sensor that attaches to any bin.”
Named after trash-loving Sesame Street character Oscar the Grouch, the technology uses a camera and ultrasonic sensor to see a person approaching a trash can. Intuitive AI’s algorithm then focuses on the items they have in their hand, predicts with very high accuracy what the piece of garbage is, and tells the person via a small screen the correct section of the bin to put it in. The system is sophisticated enough to determine the component parts of each item—so, in the case of a coffee cup, Oscar nudges an individual to throw away the sleeve, cup, and lid in different openings.
“As you’re walking up, about five, six metres away, the AI starts predicting,” Murad says. “It’s like a little child that we’ve created, with thousands and thousands of data points. We’ve had to train it very, very quickly by showing it millions of images and pieces of sensor data so that it can recognize that even though the banana you’re holding has gone brown, it’s still a banana and needs to go in the compost bin.”
On top of their creation’s obvious social impact, the pair believe that big money can be made by sorting trash more efficiently. Not only can the camera see exactly what items are entering a bin—a way for a company to save up to $10,000 on the process of auditing its waste—it can also identify the branding on the packaging. The pair hope to leverage that data to work with fast-food and coffee-shop giants to offer coupons to those who use Oscar to recycle. But the company’s biggest profits, they believe, will come from optimizing waste removal in large institutions.
“Recycling properly saves money,” Murad says. “It saves the building money, the city money, and the taxpayer money. Take airports, for example. Janitorial staff map their routine of cleaning out bins based on inbound and outbound flights. With Oscar, if it sees a big inflow of garbage or a lot of people passing by, it tells them that they need to look at those particular gates. That’s very important when you have 200 janitors who are spending 40 percent of their time cleaning out garbage bins. And it’s the same in malls or universities.”
The Vancouver company is already fielding offers. More than 60 businesses, Murad says, have expressed interest in buying or piloting Oscar, including Fortune 500 brands.
Before they make their technology available to the public, however, the pair want to make sure it is as accurate as it can be.
“More than 80 percent of waste in developed countries can be recycled on a daily basis, but only one percent is,” Murad says. “This is a way to change that.”
This article originally appeared in Straight