IOT MONITORING OF THE EXTRUSION PROCESS DURING SNACK PRODUCTION ON A TWIN SCREW EXTRUDER

Abstract

The article presents an in-depth investigation of an Internet of Things (IoT) architecture designed to enable advanced, high-resolution monitoring of the snack extrusion process on a twin-screw extruder. The proposed solution integrates the complete data lifecycle: the acquisition of process signals from sensors and inverters, preprocessing and normalization by a Programmable Logic Controller (PLC), real-time transmission through the MQTT protocol, and structured long-term storage in a Database Management System (DBMS). This approach ensures reliable data continuity and supports both operational visualization and the development of advanced analytical tools for process optimization and quality assurance. A comprehensive set of key technological indicators has been defined, reflecting the critical behavior of the extrusion system. These include multi-zone temperature profiles, pressure dynamics in the die zone, motor current and rotational frequency, precise dosing of raw materials, and cumulative energy consumption. Continuous monitoring of these parameters provides the foundation for stabilizing the process, detecting anomalies at an early stage, and improving reproducibility of product characteristics. A systematic review of scientific literature and industrial case studies published in the last three years confirms the increasing adoption of Industry 4.0 technologies within food manufacturing. Notably, the integration of lightweight MQTT-based data pipelines, the deployment of predictive maintenance strategies for electric drive systems, and the initial stages of Process Analytical Technology (PAT) implementation create the prerequisites for building a Digital Twin of the extrusion line.

Based on the conducted research, the article formulates detailed recommendations for constructing a mathematical process model, defining an economically justified but sufficiently informative sensor set, and designing an optimal data structure capable of supporting long-term industrial operation. A step-by-step roadmap is proposed, guiding enterprises from fundamental monitoring tasks toward the implementation of predictive analytics, anomaly detection algorithms, and eventually, the deployment of a full-scale Digital Twin. The literature review highlights a noticeable gap in practical, data-driven studies focused specifically on the optimization and monitoring of twin-screw snack extrusion systems. This gap underscores both the originality and the practical relevance of the presented work, as well as its potential to contribute to further research in data-centric food process control and digital transformation.