Autonomous fault monitoring for efficient multi-actuator compressed air systems : data analytics of demand-oriented parameters

Abstract

Industry 4.0 has given rise to an increased use of system data and analytics. Autonomous fault detection in pneumatic systems has also been increasingly explored in the last decade, with one of the main objectives being to reduce the energy consumption and resulting carbon footprint of such systems. Nevertheless, pneumatic fault monitoring systems tend to mainly focus on the supply-side, neglecting the demand-side. Studies covering pneumatic fault monitoring on the demand-side, perform research on simple systems, typically investigating the effects on a single actuator or a very small system. This study aimed at tackling this issue, with tests performed on an industrial multi-actuator pick-and-place setup, logging data (i.e. cycle time, flow rate and system pressure) concurrently at two locations within the system. Furthermore, different sized leaks were introduced at three distinct locations, while monitoring their impacts on the system. It was found that with the use of the average, standard deviation and impulse factor of the cycle time and the other two parameters, it was possible to identify the presence of faults on a relatively large system. For instance, the retraction time for one of the actuators reduced by 26% as one of the faults was induced. With modern industrial setups already logging the cycle time and system pressure along the demand-side, this study shows that by using existing equipment, one can develop a reliable fault monitoring system. Such information makes it possible to determine additional fault characteristics, mainly the fault's severity, type and location, thus paving the way towards smart and energy efficient compressed air systems.