Acoustic-Based Machine Condition Monitoring—Methods and Challenges

The traditional means of monitoring the health industrial systems involves use vibration and performance techniques amongst others. In these approaches, contact-type sensors, such as accelerometer, proximity probe, pressure transducer temperature transducer, are installed on machine to monitor its operational parameters. However, methods fall short when additional sensors cannot be due cost, space constraint or sensor reliability concerns. On other hand, acoustic-based technique provides an improved alternative, acoustic (e.g., microphones) can implemented quickly cheaply in various scenarios do not require physical contact with machine. collected signals contain relevant operating information about machine; yet they sensitive background noise changes condition. These challenges being addressed from applicability perspective for condition monitoring. This paper presents development methodology fault diagnostic highlights encountered analyzing sound