Fault condition indicators along the demand side of a sustainable compressed air system

Abstract

Compressed air is a vital form of energy storage and transfer in the manufacturing industry. Pneumatic system efficiency is typically low due to faults such as leakages and pressure drops which are often left unattended due to the downtime required to maintain them. The goal of this study was to identify parameters which can be used to computationally detect the presence of faults in a pneumatic pick-and-place system within an industrial automation setup. Leakages and pressure drops were systematically induced at three locations along the demand side of the system. Pneumatic data was collected from two separate locations simultaneously, and this work presents the systematic approach utilised to investigate the impacts of each fault. Indicators, namely the mean, standard deviation and impulse factor for pressure and flow rate data were identified as being able to identify the presence of faults together with their severity, type and location. This study concludes that by solely using pressure data, all three fault characteristics could be identified. These indicators can form part of data monitoring within smart machines, allowing for autonomous analyses.