Artificial intelligence is transforming plastics recycling by sorting high-density polyethylene milk bottles at speeds and scales that “would be difficult for people to match”, according to Biffa Polymers. At the company’s recycling facility in Redcar, computer vision and optical scanning analyze crushed or distorted containers on fast-moving conveyor belts, ensuring food-grade material can return to refrigerators within two weeks.
AI Sorting Technology at Biffa Polymers
The Redcar facility relies on optical and infra-red scanning alongside image analysis to identify different types of plastic. According to Managing Director James McLeary, the system is particularly useful for recognizing high-density polyethylene (HDPE) bottles that have lost their shape during disposal. While an intact milk bottle is easy to identify, crushed containers change appearance entirely. AI systems overcome this by learning to recognize various deformed shapes as valid milk bottles, as reported by the BBC.
The site has processed the equivalent of more than 10 billion plastic milk bottles in almost 20 years. Once sorted by type and colour, the material undergoes a processing cycle:
- Shredding the plastic
- Washing and separating
- Treating the material
- Turning the output into fine plastic pellets
These tiny HDPE pellets travel to other factories, where they are made into new recycled packaging, including food-grade material for milk bottles.
Workforce Impact and Plant Safety
Despite heavy automation, the Redcar facility maintains a workforce of about 200 people. McLeary noted that the technology does not replace the human labor force; instead, it frees workers from tedious tasks so they can ensure products are of the right quality. Employees still handle hard labour on the production lines.
Beyond material sorting, Biffa uses AI to enhance workplace safety. Camera networks monitor operations across the site, identifying potential risks, such as someone moving into an area where a vehicle could be operating. The company is also investigating other robotic systems that use AI to physically pick out those bottles.
Supply Chain Speed and Future Outlook
The integration of advanced machine learning has sped up the recycling process. According to McLeary, a bottle that you threw away two weeks ago may be back in somebody else’s fridge. This turnaround demonstrates how digital infrastructure can modernize traditional waste management operations without eliminating the human workforce required for hard labour on the lines.
Worth a look