This paper proposes an FIWARE-based digital twin structure to advance textile and garment manufacturing. By merging data from legacy and modern machines into a unified context, the system enables predictive maintenance and reduces downtime. FIWARE blocks convert and process real-time telemetry, ensuring interoperability and simplified analysis. A partial implementation in MTEX NS demonstrates that a continuously updated virtual replica lets technicians quickly detect anomalies, optimize resource use, and schedule proactive interventions. Future work involves transitioning to richer semantic modeling, improved identity and access controls, and machine learning-driven fault prediction. The results emphasize how open-standard ecosystems based on Digital Twin principles can modernize textile and clothing operations, bridging gaps between legacy machines, predictive insights, and evolving Industry 4.0 requirements.