Forecasting the European WEEE Collection Rate in Alignment with the SDGs: An Integrated Predictive Framework

Abstract

The management of Waste Electrical and Electronic Equipment (WEEE) remains a critical sustainability challenge across the European Union (EU). Although the WEEE Directive 2012/19/EU predates the United Nations Sustainable Development Goals (SDGs), its progressively stricter collection targets, from 45% to 65%, are closely aligned with global objectives on sustainable production and consumption. This study develops an integrated predictive framework to forecast the EU27 WEEE Collection Rate (CR) from 2022 to 2030. The framework draws on four modeling families: Statistical methods, Machine Learning (ML) algorithms, Deep Learning (DL) architectures, and selected Hybrid configurations.  Data for “Waste Collected” (COL) and “Products Put on the Market” (MKT) were obtained from Eurostat, with missing values imputed through linear interpolation validated against external socioeconomic indicators. Among all models tested, Lasso and Ridge Regression achieved the most accurate forecasts for the COL and MKT datasets, respectively. Although a Hybrid model was implemented to address non-linear residual patterns, it did not outperform the standalone Lasso model, which was retained for CR estimation. The resulting forecasts reveal a consistent downward trend in the Collection Rate, remaining below the 65% target throughout the forecast horizon. This shortfall is primarily driven by an accelerating volume of EEE placed on the market that is not matched by proportional increases in WEEE collection. The findings highlight systemic gaps in current collection mechanisms and underscore the need for enhanced policy interventions. The proposed framework offers a replicable and empirically validated tool to support evidence-based planning and regulatory monitoring in alignment with EU environmental policy and the Sustainable Development Goals.