External fault prediction maintenance modeling for fiber optics telecommunication networks

The telecommunications industry that uses fiber optics, which performs traditional reactive maintenance is no longer sufficient. The inability to predict the location and type of faults, the need for spare parts, human resources, and maintenance costs, poses serious challenges in decision-making. The Non-Homogeneous Poisson Process (NHPP) as a predictive model to measure the reliability of fiber optic networks by estimating external factors that contribute to service faults, using three covariates: number of disasters (X1), population density (X2), and number of users (X3). The best model is useful for providing solutions for resource allocation decisions (human, equipment, warehouse, spare parts, costs).