The Spartan Expert Rules Engine (SPERE) allows your subject matter experts to capture error modes as ‘if..then..else’ logic.
Once captured and implemented, expert rules automatically run against real-time data streams to trigger and alert experts when an error condition occurs.
As well as being fast to implement, expert rules can be run against the historical plant and equipment data in the APM Digital Twin to auto-generate labels.
Unsupervised Machine Learning (ML) algorithms look for underlying structure within your historical data and automatically surface clustering and association relationships within your data.
Unsupervised learning can highlight a broad range of failure modes.
Supervised ML algorithms use labels created by subject matter experts as training examples of ‘normal’ vs ‘’abnormal’ behaviour.
Labels are generated from automated expert rules and manual curation by experts.
Supervised ML algorithms are very good at catching specific failure modes that are prevalent in the training data.
To chat about how we can help you move to digital operations, complete the short form below and one of our team will email or call you back today.
Please give us a call on +44 (0)141 559 7100 or send an email to info@spartansolutions.com
