Análisis y Transferencia de Prácticas Internacionales en la Integración de Ensayos No Destructivos, Inteligencia Artificial Y Gemelos Digitales para el Mantenimiento Predictivo de Activos Industriales en Colombia

Abstract

Industrial asset management is currently grappling with the dual pressure of aging infrastructure and stringent regulatory compliance within a framework of cost optimization. While Non-Destructive Testing (NDT) remains a cornerstone for identifying structural integrity issues such as corrosion and fatigue, its traditional application—characterized by periodic inspections—fails to provide the real-time insights necessary for proactive intervention. In Colombia's mining, energy, and oil & gas sectors, unscheduled downtime accounts for productivity losses of up to 25% and drives operational budget overruns beyond 30%. This study emphasizes the strategic necessity of evolving toward predictive maintenance by integrating NDT data with digital twin technology and advanced modeling. By transforming static inspection results into dynamic inputs for virtual replicas, organizations can simulate asset behavior in real-time. This integration bridges the gap between technical diagnosis and high-level managerial decision-making, ensuring operational continuity and alignment with ISO 55000 sustainability standards.