An artificial intelligence (AI) tool that helps recycling companies better optimize their operations is the latest addition to a platform developed by Diversys Software Inc.

The Oakville, Ont.-based company’s software as a service is designed to close the data gaps that can arise when tracing materials in the recycling stream and simplify regulatory compliance. It was launched in July 2024.

In a major step forward for the platform, the company announced the addition of VERA (Virtual Environmental Recovery Assistant) in October to expand its capabilities. Now named Diversys.ai, it will add “improve” to its skillset of “track, trace, prove,” Mike Faddoul, the company’s vice-president of strategy, said in an interview with Sustainable Biz Canada.

VERA can output the work of “100 (human) data analysts crunching data over a month” in minutes, he said. That computing power can be harnessed to develop short- and long-term plans to boost the productivity of a recycling business. Diversys says operational cost reductions can be as large as 20 per cent per year.

“Our vision is a world without waste. Waste isn’t just used materials, it’s time, it’s effort, it’s work . . . it’s money,” Faddoul said.

Greater accountability in recycling

Led by CEO Roger Barlow, who co-founded Diversys in 2019, the company’s solution addresses data and accountability gaps in the recycling industry, Faddoul explained.

The sector has faced issues such as companies incinerating or dumping plastics rather than recycling the waste as expected. There have also been instances where the data collection was shown to be incomplete or flawed, leading to questionable claims and decisions. The Organisation for Economic Co-operation and Development said 22 per cent of the world’s plastic waste is mismanaged in a 2022 report.

Diversys.ai collects data from waste haulers in the field via mobile app, who will note how much waste was collected. To confirm the waste was handed off to the hauler, a person at the source of the waste will verify it was collected by taking a picture. Faddoul likened it to the way a delivery person for Uber Eats verifies the food was dropped off.

“Every single person, every hand off that’s taking place, every single person is being held accountable,” he said.

The data is fed into the Diversys.ai cloud system in real-time, creating a timeline of the delivery and naming the people who handled the waste. If there’s a discrepancy, such as the weight of the delivered waste not matching the number in the app, it will be flagged by the system.

Diversys.ai can be customized to track any post-consumer material, Faddoul said, listing examples such as plastic water bottles, aluminum cans, artificial turf and tires.

Diversys’ VERA

A major update to Diversys.ai is VERA, an AI-powered tool the company has added to enhance its platform.

Diversys began developing VERA in December 2023, Faddoul said. Unlike the large language model AI systems that have exploded in popularity in recent years, VERA learns from a company’s internal data rather than the internet.

“The data that you collect in the field that VERA is consuming and watching and learning from is not shared with anyone else,” Faddoul explained.

VERA can lighten the work burden faced by companies. It automatically produces reports to prove compliance with extended producer responsibility policies, Faddoul said. VERA can analyze a company’s data to detect any outliers in case of manipulation. It can even scan the uploaded images of waste to verify the contents and sound the alarm in case of contamination.

The AI system can also develop plans to generate financial efficiencies. As an example, Faddoul said the AI system can recognize a company is concentrating too many waste haulers in one city. In response, it will recommend diverting half of them to a neighbouring location. Once the savings are found, VERA can be asked how to best invest the funds.

Diversys was able to generate short-term operational reductions between five to 10 per cent per year with its app, Faddoul said. In the mid- to long-term, it can be between 15 to 20 per cent.

Diversys’ customers

The company’s clients are primarily producers, waste haulers, regulators and governments. The Alberta government is a Diversys client, using the platform for its beverage container recycling program.

Diversys.ai has gained significant traction in countries outside of Canada, Faddoul said. One of its largest clients to date is the California government. Diversys.ai is used by CalRecycle for its tire recycling program which processes 60 million tires per year.

The company signed an agreement earlier this year with Saudi Arabia’s government to address its beverage container and tire recycling programs, Faddoul said.

The scalability of its system is an advantage as the company looks for growth. Diversys will be announcing more government partners later in the year, Faddoul said.