AI-driven development of biodegradable flushable wet wipes from upcycled household organic waste: formulation, performance, and predictive optimization

Abstract Innovative, sustainable alternatives are needed in light of the growing environmental concerns surrounding synthetic single-use wipes and the growing amount of organic waste generated in households. This study describes the creation and refinement of biodegradable flushable wet wipes made from recycled Fruit and Vegetable Waste (FVW) combined with naturally occurring antimicrobials such orange peel oil and Coleus amboinicus . Thermochemical pulping, sheet casting, and bio-lotion impregnation were used to create the wipes, which produced a plant-based matrix with encouraging structural and sanitary qualities. The wipes thermal stability, structural integrity, and moisture permeability were validated by physicochemical characterisation using FTIR, DSC, TGA, and WVTR analysis, making them appropriate for flushable use. Significant inhibitory zones against Pseudomonas aeruginosa , Escherichia coli , and Staphylococcus aureus were found in antimicrobial testing, exceeding commercial wipes. In order to forecast important performance outcomes based on material composition, a machine learning framework was also put into place. Fibre and phenolic content emerged as the main predictors for WVTR and antibacterial efficacy, while Random Forest models produced high predictive accuracy (R ² >0.90). In order to enable scalable, low-impact bio-product creation, effort combines green chemistry, and artificial intelligence in a circular, data-driven approach to the production of eco-conscious hygiene products.