Stochastic Degradation Modeling of a Peristaltic Pump in a Biomedical Blood Autotransfusion System

Abstract

This study compares two stochastic models, Brownian Motion with Drift and the pure Wiener Process, applied to the degradation of a peristaltic pump used in a biomedical device. Failure was defined as the reduction of flow rate below a critical threshold required for safe operation. Simulations showed that the Wiener Process consistently produced higher time-to-failure (TTF) values across all scenarios, with larger differences under low and moderate variability (σ = 0.05 and σ = 0.3) and smaller differences under high variability (σ = 0.8). The observed effect results from the deterministic component in Brownian Motion with Drift, which accelerates average degradation, while the Wiener Process is driven solely by random fluctuations, allowing greater variability in failure time. Findings indicate that, in biomedical applications, selecting an appropriate degradation model is crucial for accurate lifetime predictions and predictive maintenance planning, particularly in low-variability environments. Future work should integrate stochastic degradation models with statistical monitoring techniques to improve early fault-detection accuracy and enhance the robustness of lifetime estimation under real operating conditions.