The production of recycled PET flakes plays a pivotal role in the plastics recycling industry, due to both its ecological benefits and its contribution to the circular economy. Nevertheless, the low Overall Equipment Effectiveness (OEE) observed in the production process constitutes a critical challenge. This is primarily attributed to the ingress of contaminants, blade wear, and voltage drops, all of which lead to unplanned stoppages. In response to this issue, the present study proposes an innovative solution based on the integration of Industry 4.0 technologies namely, the Internet of Things (IoT), Poka Yoke mechanisms, Predictive Maintenance, and Machine Learning. The objective is to optimize OEE and minimize the occurrence of process interruptions. Specifically, the implementation of real-time monitoring systems and predictive analytics enables the anticipation of equipment failures, thereby enhancing availability and performance. The findings demonstrate a 12% improvement in OEE, representing a substantial advance toward the global benchmark of 85%. Furthermore, the proposed methodology proves to be both scalable and adaptable to diverse industrial contexts facing similar inefficiency and downtime challenges. In conclusion, this approach not only enhances the productivity and sustainability of PET recycling processes but also establishes a replicable model for other manufacturing and recycling industries, thereby fostering the adoption of more efficient and sustainable operational practices.