Machine learning-blockchain enabled predictive framework for sustainable polystyrene waste tracking towards energy recovery

The management of non-biodegradable foam polystyrene (EPS/XPS) poses significant environmental challenges due to inefficient recycling and the generation of microplastics. This study proposes an integrated framework combining Internet of Things (IoT) sensing, machine learning (ML), and blockchain technology to transform EPS/XPS waste management into a dynamic, transparent system optimized for energy recovery. Smart bins provide real-time data streams immutably recorded on blockchain. Comparative analysis revealed PureChain as optimal, achieving 2847 transactions per second with minimal energy consumption (0.02 kWh/tx). A Long Short-Term Memory (LSTM) model demonstrated 94.8% accuracy in forecasting time-to-fill, while XGBoost offered the best balance between accuracy (92.8%) and computational efficiency. A prospective life cycle assessment indicates the framework can reduce collection vehicle kilometers by approximately 30% and enhance feedstock quality for energy recovery, delivering a net-positive environmental balance. The convergence of these technologies enables a closed-loop system that delivers significant operational efficiencies and a verifiable pathway toward a circular economy, transforming a persistent pollutant into a valuable energy resource.