Using Intelligent Edge Devices for Predictive Maintenance on Injection Molds

A considerable part of enterprises’ total expenses is dedicated to maintenance interventions. Predictive (PdM) has appeared as a solution decrease these costs; however, the necessity end-to-end solutions in deploying predictive models and fact that are often difficult interpret by practitioners hinder adoption PdM approaches. In this work, we propose flexible architecture for recommend actions. The proposed based on containerized microservices intelligent edge devices together with hybrid model which fuses generalized fault trees (GFTs) anomaly detection. Results injection molds carried out at OLI, Portuguese company, show suitable services such data preprocessing, sensor management, flow control, among others. These run near shop floor, allowing greater flexibility, they may be remotely managed customized according company’s requirements. results GFT an estimated reduction more than 63% current costs, while distribution analytics tasks reduces burden network, requiring only 0.2% cloud storage.