UK researchers use machine learning to develop “highly accurate” plastic and compostable waste sortation

14 Mar 2023 --- Researchers at University College London (UCL), UK, have developed a method to automatically sort different types of compostable and biodegradable plastics and differentiate them from conventional plastics using machine learning.  

The researchers used hyperspectral imaging (HSI) – an imaging technique that detects the invisible chemical signature of different materials while scanning them, producing a pixel-by-pixel chemical description of a sample – in a one-step process to identify various materials.

“We apply shortwave infrared in the range 950–1,730 nm to identify not just only different types of conventional plastics – PP, PET and LDPE – and compostable plastic – polylactic acid (PLA), polybutylene adipate terephthalate (PBAT) – packaging but also compostable materials – palm leaf and sugarcane-based materials – with various sizes from 50 x 50 mm to 5 x 5 mm,” the study details.

The HSI descriptions were interpreted using artificial intelligence (AI) models and material identification was made.

“The accuracy is very high and allows the technique to be feasibly used in industrial recycling and composting facilities in the future,” says Professor Mark Miodownik, corresponding author of the study.The study highlights the potential of HSI as a promising technology for real-time sorting of compostable plastics.

AI in waste sortation
The study highlights the potential of HSI as a promising technology for real-time sorting of compostable plastics and enhancing the sorting purity of plastics recycling collections and industrial composting.

According to British climate NGO WRAP, around a third of all plastic packaging on the global market leaks from collection systems, polluting the environment. Additionally, it states that plastic packaging in the UK accounts for nearly 70% of the country’s plastic waste, “making packaging the primary focus.”

“Currently, most compostable plastics are treated as a contaminant in the recycling of conventional plastics, reducing their value. Trommel and density sorting are applied to screen compost and reduce the presence of other materials. However, the level of contaminants from the current screening process is unacceptably high,” explains Miodownik.

“The advantages of compostable packaging are only realized when they are industrially composted and do not enter the environment or pollute other waste streams or the soil.”

Improving compostable plastic identification 
The team of scientists, including Nutcha Teneepanichskul,  Helen Hailes and Miodownik from UCL’s Plastic Waste Innovation Hub, tested seven types of materials and were able to identify compostable plastic with “100% accuracy.”

“The compostable plastic market worldwide is predicted to reach US$3.1 billion by 2027,” the study details.Currently, most compostable plastics are treated as a contaminant in the recycling of conventional plastics, reducing their value.

“The full environmental advantages of compostable plastic will only be realized if these plastics do not pollute other waste streams and do not enter the open unmanaged environment. HSI is a promising technology due to real-time sorting: it has high accuracy (99%), low power consumption and no additional chemicals or water are needed,” assert the UCL scientists.

The study authors share that their system can sort compostable plastics at the typical product scale (compostable spoons, forks, coffee lids) and differentiate them from identical-looking conventional plastic items with “high accuracy.”

While the model achieved perfect accuracy for all materials when the samples measured more than 10 mm by 10 mm, for sugarcane-derived or palm-leaf-based materials measuring 10 mm by 10 mm or less, the misclassification rate was 20% and 40%, respectively.

Furthermore, for pieces measuring 5 mm by 5 mm, the researchers elaborate that some materials were identified more reliably than others. For LDPE and PBAT pieces, the misclassification rate was 20% and both biomass-derived materials were misidentified at rates of 60% (sugarcane) and 80% (palm-leaf).

However, the model could identify PLA, PP and PET pieces without error, regardless of sample measurements. “Some recycling plants are interested in HSI because it is able to enhance sorting purity of plastics recycling collections and industrial composting,” the study reiterates.

Miodownik points out: “Currently, the speed of identification is too low for implementation at industrial scale. However, we can and will improve it since automatic sorting is a key technology to make compostable plastics a sustainable alternative to recycling.”

Edited by Radhika Sikaria