Soft-sensors to drive manufacturing toward clean production: LCA based on Digital Twin

This study introduces a novel methodological framework that integrates soft-sensor-based Digital Twin (DT) technology with Life Cycle Assessment (LCA), addressing data acquisition challenges and enabling dynamic environmental impact assessment. By aligning the DT standard (ISO 23247) with LCA phases (ISO 14040), the framework provides a real-time environmental analysis model. The methodology's efficacy was demonstrated through a case study on the extrusion of ten PVC materials with varying compositions and rheological properties. A soft sensor was developed to estimate the extruder motor's specific energy consumption based on screw speed, material hardness, and viscosity. Results revealed specific energy consumption ranging from 28.80 kJ/cm3 for softer PVC to 46.06 kJ/cm3 for harder PVC at 120 rpm screw speed. The framework facilitated real-time environmental impact quantification, showing Global Warming Potential (GWP100a) between 0.59 and 0.95 kgCO2eq per gram of extruded PVC, contingent on material properties and operating conditions. Key outcomes include a real-time environmental impact model with R2adj = 0.84 and sigma = +/- 2.20 kJ/cm3, a potential GWP100a reduction of up to 16.4% through operating condition optimization, and up to 48.7% through eco-design-driven material selection. This research bridges Industry 4.0 technologies with LCA, offering a dynamic, real-time approach to assess and optimize environmental impacts. It contributes to the transition toward Industry 5.0, paving the way for more sustainable manufacturing processes.