Smart Reliability Estimation Via ANN-ABC Optimization: A Novel Approach to Inverse Weibull Process Under NHPP

Abstract

In this work, a new intelligent hybrid estimation framework is proposed to model the time-dependent failure behavior of a repairable system. The model adopts an Artificial Neural Network by applying the Artificial Bee Colony (ANN-ABC) algorithm. It also combines the Inverse Weibull Process (IWP) with the Nonhomogeneous Poisson Process (NHPP) framework. The study compared the traditional Maximum Likelihood Estimate (MLE) model with the proposed hybrid algorithm using simulation. The results showed that the ANN-ABC model has lower Root Mean Square Error (RMSE) and Bayesian Information Criterion (BIC) when a moderate-to-large dataset is used. In addition, the proposed model was validated using real clinical data from 299 heart failure patients to predict the survival likelihood. The results validate that the proposed hybrid intelligent methods can be used to support AI-driven reliability modelling for complex data driven systems such as monitoring, predictive maintenance, and intelligent healthcare systems due to their superior support in the precise estimation of failure intensities. This contributes to SDG 3 (Good Health and Well-Being) and SDG 9 (Industry, Innovation and Infrastructure) by enabling more reliable intelligent healthcare and resilient cyber-physical infrastructures. © 2026 NSP Natural Sciences Publishing Cor.