Statistical Process Control with Widget Working View

Our Numerical Analysis team has been busy behind the scenes working with the developers to bring you Statistical Process Control  and Anomaly Detection – it’s right there, at your finger tips in the Widget Working View.

When you next load up Widget Working View you will see an Analytics tab has been added. This will enable you to analyse your data right there in the browser, without needing any other appplication. The analytics are broadly applicable statistical techniques with good reason; you can use them for anything.

They’re great for both project and process data; identifying problematic trends in projects, alerting when quality indicators fall outside limits or identifying repeated issues in production.

Anomaly Detection for time series charts uses a suite of algorithms combined with quorum voting to highlight anomalous behaviour in your data. You zoom in and investigate these anomalies using the handy tools.


Statistical Process Control (SPC) rules have been applied to the time series charts along with Anomaly Detection. These rules can be configured by the user  so if any rules are broken a mark will be placed on the chart to highlight where that rule has been broken. We have currently implemented four rules which can be set.

Out of Bounds: This highlights any points that are over or under a set band. This could be used to show when a temperature or pressure is outside of the controlled limits.


Consecutive Points: This highlights the number of consecutive points above and below a band. One use for this would be to indicate that there has been a step change in process so you can then check for feed rate changes or process changes upstream.


Upward / Downward Consecutive Points: These show a number of points where the data has trended in a certain direction. As statistically it’s unlikely to get a long sequence of points in a given direction, to do so may indicate a problem, e.g., that pressure is building.
This is useful even where correction occurs as customers have found it can indicate a leak or fouling which is then corrected by process control systems.


We have a sandbox full of charts and analysis tools which could revolutionise how you see your data. For more information on how Sabisu Analysis can transform your data contact us for a live demo.

Contact us

We’re always interested in hearing from you with any comments or suggestions, feel free to get in touch.

Start typing and press Enter to search