Recoup battles UK littering with new AI-based data collection project

09 Aug 2023 --- Recoup is launching an initiative to obtain better litter data in the UK. The Litter Composition & Pathways Project (LCPP) aims to provide data for making informed decisions and developing effective environmental policies.

Proposed reforms aimed at reducing litter rely on accurate and robust data sets. A comprehensive understanding of litter composition is essential to assess the impact of these reforms and determine the most effective strategies for curbing litter pollution, stresses the charity and leading authority in the resources and waste sector.

But current litter projects often face limitations in data collection methods, leading to a lack of consistency and reliability. Studies often rely on citizen science, which only sometimes provides reliable data, warns Recoup.

The project aims to review, challenge and provide robust evidence regarding litter composition. To achieve this, Recoup aims to develop a standardized, evidence-based methodology for collecting and analyzing litter, creating a comprehensive data set. Peer reviewing of the methodology will ensure the reliability and validity of the collected data.

“There is a real push at the moment from the regulators, industry and the public to tackle environmental matters across the board and with litter being one of the primary and most visually obvious challenges that we need to tackle, this project is set to provide the groundwork for future work that can seek to address the issues of litter in the UK, and hopefully, in time, further afield,” Tom McBeth, policy and infrastructure manager at Recoup, tells Packaging Insights.

The Recoup project aims to review, challenge and provide evidence regarding litter composition.Overcoming challenges with AI
Recoup’s LCPP project engages stakeholders and partners to collect and assess this material, producing a transparent dataset that can advise future practice.

“Of course, it comes with challenges. The range of material types, formats, conditions and locations add significant variables to being able to make assumptions, and appropriate volumes from different areas will need to be assessed to generate a reflective dataset for the UK,” explains McBeth. 

He continues by saying discussions are taking place regarding using AI and developing technologies across the whole waste management sector.

“We are seeing great efforts and interest in AI’s use in sorting and identification at material recovery facilities and similar, but litter is a completely different beast. This is due to the significant variables in the material and how they are likely to be presented – whether that be contaminated by the environment it has been in, broken or damaged.”

“However, with the right dataset developed as part of the project, we are hoping to be able to develop and educate these systems, expand the data within them to identify these types of materials, and in time be able to improve the effectiveness and accuracy of these technologies,” asserts McBeth. 

Collaborations ahead
Although the LCPP is still in its early stages, Recoup says the project already has a draft methodology for litter collection, and sorting is ready for testing and peer reviewing. The initial phase of analyzing litter using AI technology has begun.

The charity is establishing a “Litter Lab” to further enhance the project’s capabilities at its offices, where advanced litter analysis will be conducted. Recoup plans to collaborate with local litter-picking groups to collect waste for analysis. 

Once the methodology is finalized and the data is collected, it will be shared with the UK government, devolved administrations and key stakeholders to inform and advise policy decisions. The evidence-based insights gained from this project are said to aid in developing targeted interventions to combat litter pollution.

“Initial engagements have been on a local level, but the project will be scaled up to cover the whole UK. It’s important that assumptions aren’t made when it comes to litter – and it’s clearly unfair to base volumes and composition of litter found in densely populated areas to the extremely rural,” concludes McBeth. 

By Natalie Schwertheim