Faraday Predictive

Contact: Geoff Walker <geoff.walker@faradaypredictive.com> and Will Boulton <will.boulton@artesis.co.uk>

Predictive Maintenance of Industrial Equipment

Faraday Predictive is a small local technology company which uses sensor data from electric motors to diagnose faults in rotating equipment (e.g. fans, pumps, compressors), and provide advice to clients about maintenance of their equipment. Useful advice contributes to the smooth running of this industrial equipment, potentially saving customers millions of pounds in some cases.

One of our customers operates in the gas industry. We have an expanding dataset (still in use) of thousands of records from 72 identical gas compressors, some of which have shown problems, that could be used to train machine learning models, rather than relying on expert analysis of each compressor. Your task is to deliver a system that can easily be used both by site engineers to alert them to problems, and allow them to provide feedback (alerting your system to when and where maintenance has occurred), and site managers, to schedule maintenance based on your models of machine health.

Potentially a project related to the company's work in condition monitoring of rotating machinery, for example using machine learning for failure prediction. However, this would have to be integrated into a larger product concept.

Alternatively, perhaps adapt The Headless Bicycle?