This study presents a reliability analysis of an ice cream manufacturing facility based on operational failure data. The research develops a ten-component system model that captures plant dynamics and organizes it into three subsystems to reduce computational complexity. The study derives key reliability parameters and constructs a state transition diagram to represent system behavior across multiple operating states. The analysis applies an Artificial Neural Network (ANN) to address the limitations of conventional analytical methods in modeling complex nonlinear systems. The ANN provides strong predictive performance and manages uncertainty within the operational environment. The proposed framework estimates system reliability with high precision and supports maintenance planning and cost optimization. Numerical simulations validate the effectiveness of the ANN-based model. The results demonstrate improved prediction accuracy and greater computational efficiency compared to traditional approaches. State probability deviations are evaluated over a 24-hour period. The up-state probability increases from 95% to 96.02% during the useful life period and from 95% to 96.03% during the wear-out period. The findings confirm that the proposed method enhances reliability prediction, improves maintenance scheduling, and supports cost control and equipment design optimization in ice cream production systems.
Tewatia et al. (Wed,) studied this question.