Mechatronics students tackle trash, secure spot in NextAI program
Each year, the world’s urban populations generate around 1.3 billion tons of waste. That’s expected to soar to four billion tons by 2100.
Despite this startling statistic, only two per cent of garbage is currently recycled or composted. The rest ends up in landfills, where it may pose a potential risk to wildlife and the environment.
“Ultimately, recycling is simply not an intuitive process at the moment,” says SFU mechatronics master’s student Hassan Murad.
“There are many variables when it comes to packaging, and recycling rules change from city to city, which creates even more confusion.”
It’s a familiar problem in many households, but perhaps not for much longer.
Murad and SFU mechatronics undergraduate student Vivek Vyas plan to develop a “smart” bin that will use artificial intelligence techniques to automatically sort recycling from trash.
The pair’s company, Intuitive Robotics, was one of only 19 startups selected for the NextAI, a global artificial intelligence (AI) tech accelerator program, which runs from May to September at the University of Toronto.
The intensive five-month program brings together promising entrepreneurs and students from across Canada and around the world to develop AI solutions for commercialization.
“The program allows you to share ideas and get feedback on your progress really quickly, so you can move forward much faster,” says Vyas.
Teams receive mentorship from professors at top North American universities and training from leading technology companies. Up to $150,000 is also available to further develop the technology, depending on their progress.
It’s a dream come true for Murad and Vyas, who launched their company just seven months ago, after Murad graduated from his undergraduate degree in mechatronics at SFU.
The duo immediately set their sights on developing a bin that would take the guesswork out of recycling. The goal: to improve recycling rates and reduce the waste and pollution associated with ever-growing landfills.
But they faced a technical challenge—the bin’s image-recognition system would need to be able to identify any item of trash from various angles and at different stages of decomposition.
“We thought, how does a child learn to recognize a banana peel, when it can be yellow or brown? By seeing numerous examples. That’s when it occurred to us to use deep-learning methods,” says Murad, referring to a machine-learning technique inspired by the structure and function of the human brain.
“That way, we could train the system by showing it thousands of example images, so it would ‘learn’ to correctly identify different items and automatically select the appropriate bin.”
“That was the turning point – right after that, we saw the call for the NextAI program,” says Vyas.
For the next five months, Murad and Vyas will be fully immersed in the field artificial intelligence – gaining knowledge from top experts including professors from Harvard, NYU and MIT.
In addition, they will learn how to further develop their product, currently at the prototype stage, and grow their business with mentorship from industry partners such as Apple, Google and IBM.
The program culminates in September in a final pitch to investors, venture capitalists, business leaders and others.
“This week, we’re in back-to-back meetings—we’ve secured our first round of funding for $10,000 and are now headed towards the next target of $40,000,” says Murad.
“The feeling of running an actual business has started to settle in.”
With such a frenetic schedule, shared vision is key to maintaining momentum. The pair hopes the “smart” bin will be the first in a range of products aimed at moving closer to a zero waste world.
“We want our products to help educate people about recycling and make the process more intuitive – that vision is so important,” says Murad.
Vyas agrees, and adds:
“We’re developing a socially responsible product that could make a huge impact in the world.”