A Measurement Frequency Estimation Method for Failure Prognosis of an Automated Tire Condition Monitoring System

The ongoing digitalization allows operators and manufacturers to constantly gain new insights about their asset’s performance and degradation status. This information could potentially help to reduce operating and maintenance costs. Although significant amount of research has been spent in determining Remaining Useful Lifetimes (RUL) of various systems, these efforts often implicitly assume an unrestricted availability of measurement data. However, the amount of acquired data significantly drives the necessary investment cost or is sometimes even impossible to obtain in required frequencies in reality. In this paper, we will investigate the changes of the precision for the RUL prognosis on the example of a Tire Pressure Indication System (TPIS). After a possible layout with sensor requirements for a fully automated condition monitoring system has been developed in theory, we describe necessary data cleansing steps to account for environmental impacts on the system’s performance and to derive the system’s health status. With the help of a Monte Carlo (MC) simulation, we evaluate the system’s sensitivity towards changes in precision of the RUL for different measurement frequencies, prognostic models, and parameter settings. The results allow an estimation of the minimum pressure measurement frequency for a fully automated TPIS in order to obtain the required prognostic performance and to maximize cost efficiency.