Despite good intentions to support the environment, many companies still send recyclable items to landfills.
The world is producing almost twice as much plastic waste as it did 20 years ago, yet only 9% of it is being properly recycled, according to the Organization for Economic Cooperation and Development.
Glacier, a San Francisco-based company founded by Rebecca Hu, a former Bain consultant, and Areeb Malik, a former Meta software engineer, aims to change this.
Using AI-enabled robots, Glacier helps companies efficiently sort through trash. Their robots can process 45 items per minute, identifying over 30 different materials to find recyclable items that humans might miss.
“The robot’s arms then pick up and sort items accurately.” This technology helps prevent over 10 million items per year from ending up in landfills, according to Glacier estimates.
Hu recommends starting with the “last chance line” or residue line, where items that are supposed to be trash are sorted. Often, recyclable items end up on this line because sorting is so challenging.
Although public awareness and action on recycling are improving, there’s still room for growth. The current recycling rate in the U.S. is around 32%, and the Environmental Protection Agency aims to reach 50% by the end of the decade.
In March, Glacier received investment from Amazon’s Climate Pledge Fund, marking the company as the second woman-led climate tech venture supported by Amazon, which has committed $53 million to women-led climate tech companies.
For Hu, this venture is personal. “It’s the first time I’ve felt an idea was so compelling that I was willing to take the risk of an early-stage startup,” she says. Growing up as a first-generation American in the Chicago suburbs, her Chinese immigrant family was always resourceful, reusing materials whenever possible.
“This reduce, reuse, recycle mantra was constant at home, in contrast to the consumerist society I saw around me,” she recalls.
With Amazon’s backing and interest from companies wanting to improve their environmental impact, Glacier aims to expand its technology and enhance its AI capabilities. They are developing computer vision tech to identify a wider range of recyclable items, from general cardboard to specific items like cat food cans.
“We have access to hundreds of millions of images of recycled items for training and validation,” Hu says. “The challenge now is not collecting these images, but figuring out which items to detect.
We collaborate closely with recycling facilities to understand where the most value lies.”
Despite the success of Glacier’s robots, Hu believes humans remain crucial to the process. “Whether we will ever have a fully automated recycling facility is still uncertain,” she says. “However, there’s a significant labor shortage in this industry.
By having robots handle the dull, dirty, dangerous tasks, the existing workforce can focus on maintenance, repair, and monitoring—roles that better utilize human skills.”
