Published: April 6, 2022 By

Each year, over 90 million tons of recyclable material end up in United States’ landfills. 

The inefficiency of the recycling system in the U.S. is largely due to contamination, or the disposal of non-recyclable materials into the recycling stream. It is estimated that, on average, 1 in every 4 items in recycling bins is a non-recyclable item. Historically, most of the U.S’ recycling has been sent overseas to be disposed. However, our contamination issue has become so problematic that China, our primary recycler, has begun to reject our recycling. The rejection of our recycling by foreign countries signifies a need for change in the way recycling is sorted in the U.S.. One potential solution to our waste problem is the use of machines, Artificial Intelligence (AI), and robotics to sort recycling. Environmental innovators have already begun implementing these technologies in select recycling facilities across the country. Widespread application of these technologies could significantly reduce contamination in the recycling system and be a huge step forward in fixing the United State’s waste issue.

Optical Sorting

Recycling sorting has been mechanized for decades in the United States. Optical sorters are one of the most sophisticated, non-AI, sorting machines. Optical sorters use the properties of light waves from different materials to weed contamination out of recyclable material. There are three key steps to this process. First, the machines move the recyclables under a bright light. This causes the different materials to absorb and reflect light waves at different levels, creating a distinct light “signature.” Second, the Optical Sorter assesses these light signatures with a spectrometer. Finally, once a targeted light signature is identified, a computer calculates the location of the item on the belt and a blast of air removes it from the rest of the assessed material. This technology is capable of identifying plastics, glass, wood, paper, and cardboard. It is also capable of finding more specific types and colors of plastics and paper. This machinery allows for more efficient sorting than what can be done with manual labor.

Check out this slow motion video of the plastics optical sorter at the Boulder County Recycling Center:

AI and AMP Robotics

Optical sorters are efficient and largely successful at removing contamination, but even better recycling technologies are beginning to be implemented. These new technologies rely on the use of AI, or artificial intelligence. AI has the ability to perform complex tasks that would typically require human intelligence. One of the most widely used AI recycling sorters is designed by AMP Robotics, a Colorado Front Range-based company created by CU Alumni Matanya Horowitz. AMP’s AI, AMP Neuron, uses computers to map millions of images of material streams to identify different types of papers, plastics, and metals. This powers AMP Cortex, the robotics system that performs the physical task of sorting the recycled material. The AMP Cortex robot has three arms that allow for a large range of motion and fast picking and placement. Together, AMP’s AI and robotics system can quickly sort through all nearly all recyclables and contaminants found in single-stream recycling.  

Trash Bot and other AI Robotics Companies

AMP Robotics is just one of many companies using AI and robotics to revolutionize recycling in the U.S. ZenRobotics is the first company to apply these technologies to recycling. This company's robotics are capable of picking up as many as 6,000 pieces of waste in one hour. CleanRobotics uses AI to sort recycling on a smaller scale. They created the TrashBot, a disposal bin that sorts waste into trash and recycling on the spot. Everest Labs and Greyparrot have also created AI waste recognition software that provides analytics on sorted materials.

AI Benefits in Recycling

There are many pros to using AI to sort recycling. These computers are able to store and process data quicker than humans can. The stored data can then be used by facility managers or even made publicly available to improve output. Even the machines themselves can learn from the information they store, allowing for a continuously more accurate sort. AI will also reduce our reliance on manual sorting. Sorting is dangerous. Workers can be exposed to harmful chemicals, respiratory hazards, or biological substances that were incorrectly disposed of, injured by malfunctioning machines, or left injured from the repetitive motion of sorting. By letting machines take on the sorting responsibility, we can utilize those workers for other roles in the recycling industry. Currently, more than 1.1 million people are employed in the recycling and reuse industries. Only 23,000 of these people are employed in recycling facilities, with an even smaller percentage being manual sorters. Ultimately, utilizing AI will not significantly impact job availability.

    The use of AI in the United State’s recycling industry is one step towards having a more efficient recycling system. It will help us to reduce contamination, make better use of the products we have already created, and reduce the environmental footprint of the United States. To learn more about this topic please check out the article linked here.