Meet Oscar, the Canadian-built, AI-driven recycling system helping governments go green

Business

Standing in the middle of a cafeteria at Global Affairs, Soren Antosz was testing Oscar, a new, high-tech tool popping up in federal government buildings.

Antosz holds up a paper napkin. That’s compostable, flashes the message on Oscar’s screen.

Next is a plastic bottle. Oscar tells him to empty the liquid then drop it into the recycling section.

With a bit of a smile, Antosz holds up his cellphone. “Put it back in your pocket,” Oscar jokes.

As the federal government begins to embrace artificial intelligence, the Oscar Sort recycling station is one of the more unconventional ways the government is using AI.  

The Oscar station looks like a standard office recycling system, but it has a scanner that checks the item which someone wants to discard, then nudges the user to deposit their garbage or recyclable in the right bin.

Currently, Oscar stations are located in 10 federal government buildings in the National Capital region. While the stations in two Global Affairs buildings are part of a pilot project that began a year ago, the stations in eight buildings operated by Public Services and Procurement Canada passed the pilot project stage in March 2022 and are now part of its regular recycling system.

WATCH | Soren Antosz shows how the Oscar Sort recycling station works:

Deputy director of EcoGAC at Global Affairs Canada Soren Antosz explains how the department is using artificial intelligence to help it divert trash from landfills and into recycling programs.

Federal officials are trying to reduce the amount of garbage from government buildings that ends up in landfills. In the case of Global Affairs, the goal is to increase the amount of garbage diverted to recycling from 54.6 per cent in 2022/2023 to 75 per cent by 2027, according to its sustainable development strategy.

AI tool built by B.C.-based company

Oscar is the brainchild of B.C-based start-up Intuitive AI, co-founded by Hassan Murad who immigrated to Canada from Pakistan, and Vivek Vyas who came from India.

Growing up in their home countries, both saw first-hand the problem of waste and recyclables shipped from the West polluting land and waterways.

The two, who met while studying robotics at Simon Fraser University, decided to use artificial intelligence to do something about that problem.

“We asked ourselves a simple question, which was: what is the biggest problem that exists at the moment that we care most about solving?” Murad said in an interview with CBC News. “And we saw that a lot of this waste was ending up in landfills or illegally being shipped and ending up in oceans.”

People in masks clean garbage.
Volunteers clear plastic and other garbage littered on the shores of the Arabian Sea on World Environment Day, in Mumbai, India, in 2022. (Rafiq Maqbool/The Associated Pres)

Founded in 2017, hundreds of Intuitive AI’s Oscar Sort systems are now located in 20 countries, said Murad.

While the name Oscar evokes images of Oscar the Grouch for many Canadians, the popular character from Sesame Street who lives in a trash can, Murad said the name actually started as a short form for what Oscar is – an object scanner and recognition tool.

A recycling bin that learns on the go

What separates Oscar from run of the mill recycling stations is artificial intelligence. A user shows the object they are about throw out to the scanner and the Oscar station will tell them where it should go. The more often a station is used, the more it learns which objects should go in which bin.

Oscar stations can be programmed for each client and location, said Murad. For example, in the National Capital region, Quebec and Ontario have different recycling streams, so the stations in federal government buildings in Quebec give different messages than those located in Ontario.

Hassan Murad
Hassan Murad is co-founder of Intuitive AI which developed the Oscar Sort recycling system. (Screen capture)

They can be set up to give users a reward for throwing something into the right bin or a “grouch” if they get it wrong with a red screen and an X. They can even be programmed to reward a user with a QR code that can give them a discount on a product – something the government departments aren’t currently doing.

The stations at Global Affairs include a recycling trivia game where users can answer questions by showing a thumbs up or a thumbs down. 

Antosz, deputy director for sustainable government operations at Global Affairs, said the reaction from employees has been positive. 

“People find it’s kind of a fun experience,” he said. “Because it’s interactive, people kind of get a kick out of it.”

Antosz said they see more people using the Oscar stations each month. In October alone, there were 15,000 disposals at one station.

Antosz said the stations are also providing the department with data that can identify problems such as people throwing used paper napkins in the paper bin rather than compost.

Antosz said they can see “how many disposals into the recycling centre were correct versus how many were incorrect and where we are seeing contamination in our waste streams. So, we can then use that data to target our messaging.”

Public Services and Procurement Canada is also seeing an increase in the use of the Oscar stations in its buildings, said Jérôme Gagnon, who was involved in the department’s initial pilot project at the C.D. Howe building in downtown Ottawa. 

“We have (had) a great reaction since the launch of the project pilot; we have seen a slow but steady improvement in terms of recycling and waste diversion at all four Oscar stations at C.D. Howe.”

Now, the government is starting to get calls from other organizations interested in the technology, said Antosz.

“When they start to hear about what we’re doing, absolutely they want to come by and see it,” he said. “They want to try it. There are quite a few cases where you’re starting to see new Oscar units around town as people start to learn about this technology.”

Leave a Reply

Your email address will not be published. Required fields are marked *