Measure recyclable resources in the waste industry to meet the foundational industries demands for recycled material using AI for tracking SKU

点击翻译按钮获取中文摘要

The foundation industries (plastics and chemical industry, paper, metal etc.) use large volumes of virgin feedstock in the manufacture of packaging. Substitution of part of this feedstock by recycled waste packaging is reducing this demand, save money, the energy used in processing by 75%, and greenhouse gas emissions by up to 2148 kgCO2e/t. Accelerating this shift is pressure towards sustainable packaging that contains sufficient recycled content, and regulation; the UK government is introducing a tax on packaging that doesn't contain 30% or more recycled content. Thus, the demand for recycled material is inexorably increasing. However, the low recycling rate (predicted to be 13% by 2040, while quantity of plastic waste will double) due to inefficient, inaccurate (manual) auditing and processing, driven by a lack of data, thus inhibiting the effective sorting of waste, limits our ability to feed the industry with enough reclaimed material. There is an urgent need to decrease the stress on new resources by leveraging the value of waste. To achieve this, transparency/accountability across the product life cycle need to be established and the only way to do it is to digitise the industry, providing it with the necessary data. Across all sorts of materials - plastics, paper, glass, metals and others - recycling systems spanning producers, recyclers, and foundational industries could be far more efficient when powered by data:more effective sorting, data-driven optimisation decisions, enforcement of regulation, better policy decisions guided by data. This project aims to build on Greyparrot's existing AI vision technology which is a novel, affordable system that uses state-of-the-art approaches in AI-computer vision to accurately identify and classify in real-time product packaging/waste at a brand/SKU level, recognising and tracking these all along the sorting process across different sorting lines, informing on whether each one is correctly being recycled or not. Although securing strong interest (with ongoing trials in MRFs) the technology is limited in terms of number (100s) and type (plastic) of SKUs being currently recognised. This project will significantly scale-up this solution through hardware and software development allowing for the recognition of a representative market share (70% of PET/HDPE main SKUs, main non-plastic packaging ones), helping to hold producers accountable for poor packaging design, enabling intelligent sorting and policy makers/waste managers to take informed decisions to adapt and improve regulations and processes, increasing the percentage of packaging being effectively recycled and going back into the manufacturing process in foundational industry.Thanks to this project and through larger scale hardware deployment and data gathering and processing, Greyparrot will increase the speed of product development, moving toward a shorter timeline for the product to reach the market and start having an impact on the foundational industry and other stakeholders.