A COMPREHENSIVE APPROACH TO TEXTILE PLANT DIGITAL TWINS VIA CYBER-PHYSICAL SYSTEMS

This work explores the practical implementation of Cyber-Physical Systems (CPS) within an industrial context, with a specific focus on enhancing data acquisition, automation, and monitoring processes to create a Textile Plant Digital Twin. The research is conducted in Medellín, Colombia, in collaboration with Leonisa, a prominent Colombian company specializing in underwear and women's clothing. The primary goal of this work is to improve the efficiency and reliability of industrial operations by applying CPS principles, a DevOps approach. The study successfully develops an effective system for data acquisition, incorporating various methodologies such as Node-RED services, Python applications, RFID tag tracking, and water turbidity measurement. Furthermore, the adoption of DevOps principles, emphasizing Infrastructure as Code (IaC) with the use of Ansible, automates server and edge devices provisioning, configuration management, and periodic updates, reducing manual intervention and enhancing system reliability. The importance of monitoring and visualization within the CPS framework is underscored by the seamless integration of Prometheus Node Exporter and Grafana. This integration facilitates extensive real-time system monitoring and proactive alerting mechanisms, particularly for the health status of numerous devices, like servers and edge computing devices, such as Raspberry Pis and other IoT devices. Additionally, the research delves into well-established CPS architecture models, including the 5C architecture, Reference Architecture Model Industry 4.0 (RAMI 4.0), and the Industrial Internet Reference Architecture (IIRA), providing guidelines for the integration of various technologies, such as digital twins, IIoT, and cloud computing, while addressing challenges in interoperability and security. The research also categorizes industrial variables into four key groups: Process, Transactional, Reliability-Centered Maintenance (RCM), and Auxiliary variables, crucial for optimizing production, product quality, and operational efficiency. The implications of this work, measured by LoA, extend beyond the textile industry, offering valuable insights for CPS implementation in various industrial domains. This project lays the foundation for the integration of CPSs, digital twins, and DevOps practices in the industrial landscape, bringing new levels of efficiency, productivity, and sustainability in the company. The continuous assessment on Levels of Automation (LoA), employing the Dynamo methodology, reveals significant enhancements across multiple facets of industrial operations. The integration of Node-RED applications, Python developments, RFID tracking, and Turbidity Measurement demonstrates a noteworthy increase in the LoA, signifying improved automation and efficiency in data acquisition processes. The detailed assessment across various stages, including system configuration, data acquisition, data transformation, data transmission, error handling, maintenance and updates, and monitoring and reporting, highlights the specific areas of advancement within each technology implementation.