10180459RecycLoop AI: Feasibility to Raise Recycled Content and Cut Scrap in UK Plastic PackagingClosedCollaborative R&DInnovate UKFrontiermind AI Ltd will deliver a two-month feasibility project to test **RecycLoop AI**, a practical decision-support tool that helps UK plastic packaging manufacturers **use more recycled material and generate less scrap** while keeping products within specification. The project addresses a clear industry need: manufacturers want to reduce virgin plastic use and waste, but changing resin blends and process set-points can risk quality failures. RecycLoop AI is designed to make those choices safer and easier. The project will focus on common packaging processes such as film extrusion, thermoforming or injection moulding. Using a short, anonymised production dataset and a small number of benchtop checks, Frontiermind will build and test a **data-light optimisation prototype**. The prototype will recommend process set-points and blend ratios that maintain required properties (for example, strength, clarity and dimensions), while **reducing virgin resin per unit** and **cutting rework/scrap**. Recommendations will be operator-readable and bounded by guardrails, so they can be reviewed before any trial on a live line. Key outputs will include: (1) baseline resource-use metrics for a representative packaging step (virgin plastic per 1,000 units, recycled content, scrap rate), (2) an offline prototype of RecycLoop AI, and (3) an evidence pack summarising results and the route to an industrial line trial after the grant. All work will be carried out in the UK, with UK exploitation of results, and without disclosing any commercially sensitive information. If successful, the project will demonstrate a **measurable reduction in material use and waste** in plastic packaging production. This supports the UK's circular-economy ambitions by making it easier for manufacturers to increase recycled content, improve yield and resilience, and prepare for evolving reporting and compliance requirements.