Trouble Detector¶

Client: Rob Humphris, 8power Ltd. rob@8power.com

There are a large number of vibration analysis systems on the market, the majority of which have attempted to automate vibration fault analysis without significant success. 8power have been focusing to try and improve the resilience of the water industry and have a tried and tested IoT Vibration monitoring system that has been collating hourly measurements across a wide estate of assets for the last 5 years, with a number of examples of failures and incidents during that time.

Currently 8power have a range of manual assessments to identify the causes of excessive vibration, but an automated method has not been successfully identified. At first glance this does not appear to be too difficult a problem to solve, but all the major companies who have developed vibration sensors have tried to do so and have failed. The complexity of the environment in which rotating equipment operates means that even a brand new asset immediately upon installation adopts characteristics of an older asset. Traditional assessments versus ISO standards for rotating equipment vibration are thus ineffective – almost all assets appear to be performing poorly against those standards. It is very possible for a human operator to undertake the analysis required, but comparison of vibration signature data and tracking of metrics over time should be capable of delivering the same assessments. Industry has tried to do this with point assessment and that is not possible without ultra-high resolution data, which can be obtained with dedicated measuring equipment but not by using condition monitoring via connected wireless sensors.

The goal of this project is to develop a Trouble Detector that will replace the manual process we currently employ. Initially this may start as a way of assisting the engineer with their diagnosis, and ideally would extend to guiding them to an assessment.