OSIsoft 2018 Hackathon - Servelec Controls Team 2 - SQC Chart

Blog Post created by Michael.Nelson on Apr 15, 2018

Who was at the gig?

Nick Alderton , Lavanya Jayakkumar , Michael Nelson .



We arrived at home base bright and early, Costa Mochas in hand and we asked ourselves, what do the people want from PI Vision? Well we checked and it was unanimous!

  • "We have been waiting for this ever since PI visializations became webbased."
  • "This is key functionality for some of my customers and I would like to add my vote for this to be considered."
  • "Agree with others that this would be a very valuable addition to Pi Vision, complementing the Pi AF functionality."

Well who are we to turn down a request from the people? Servelec Controls Team 2 is happy to present the Servelec SQC (Statistical Quality Control) Chart Symbol!


What is an SQC Chart?

SQC charts are used to identify instances of unnatural fluctuation so that causes can be assigned and corrected. Control charts are employed by a wide range of industries and agencies as a means to monitor and stimulate improvements in many types of processes.
The SQC chart generally made up of 3 key sections, an Information pane, a histogram and a chart. The chart uses statistical methods to establish baselines (usually called control limits) using data that represent ‘good operating conditions.’ Statistically-derived control limits outline the potential behaviour of a process, and permit differentiation between random fluctuations of data and true process shifts. A control limit can be defined by a specific number (centre line+/- 3 standard deviations). Utilizing these limits the chart can be used to detect process variations that have definable causes by highlighting instances where process data runs outside the control limits.


What can an SQC Symbol in PI Vision can do for you?

PI Vision a great tool that offers an opportunity for people to quickly and easily access and create displays. However at this time there is no functionality for SQC Charts to be visualised.

A user utilising our SQC chart would be able to look for changes in performance against known limits over a set period based on the Standard Deviation and Process Potential.

With the power of PI Vision a user could take this information anywhere, allowing them to be more proactive and drive better business decisions, acting rapidly on the SQC Chart information at hand.


SQC Symbol Configuration



SQC Chart Symbol

The SQC chart has been developed entirely within angular 5 and is hosted within the OSIsoft PI Vision environment. The SQC chart is made up for 3 key sections, an Information pane, a histogram and a chart. Combined, these sections allow proactive monitoring and in time early notification of potential changes in condition.
Whilst we have opted to allow simple replication of key components within the SQC chart using PI Asset Framework, a number of sections within the SQC chart are configured through code presenting a number of mathematical challenges along the way.
Examples of this are the histogram, including the bell curve and the chart which are presented to the user upon adding the AF configuration to the SQC Symbol.

Process Potential is a key figure in our SQC Chart and is being calculated as follows: (Cpk or Process Capability Index and is calculated as Min((Mean - LSL)/(3*SD), (USL - Mean)/(3*SD)) ).


PI Asset Framework

Whilst it is noted that key functionality such as Standard Deviation and Process Potential are worked out through code, there are key attributes that must first be set in PI AF.
The following template details the key parameters required in the configuration of the SQC PI Vision symbol:

image-5 jive-image

Configuring these attributes and applying it the form of a Template in AF allows users to easily store and modify the necessary configuration and output the configuration to the PI Vision SQC Chart Symbol.
In the following example we have utilised an AF template in order to quickly complete this action for our Fabian asset.

image-6 jive-image

Hackathon Business Example

In our example we have three assets Fabian, Irwin and Marti and we are monitoring the performance of each assets oil production. We are trying to ensure that our Process as close to the centre line as possible. We can see the current Process Potential in our charts and are aiming for as close to 3 as possible on a scale of 0 to 3.

In the following screenshot we can see that Fabian oil production was not up to scratch for this day. Whilst the Sister asset Irwin was far more on point thanks to its more consistent production flow.

image-2 jive-image


Video Demo


GitHub Link