Predictive Performance Measurement System Examples

The following predictive performance measurement system examples demonstrate an enhancement to traditional business performance management system reporting. Gartner states organizations will increase profitability through predictive measurements. The described improvement to traditional reporting provides an effective means to accomplish this predictive performance reporting objective, which can lead to improved business profitability.

The reporting that will initially be highlighted in this blog addresses data from an equipment effectiveness (OEE) example report-out; however, the techniques are applicable to a variety of situations that have a similar reporting response as part of a performance measurement system.

After completing this initial illustration, this blog will then reference other blogs for many more predictive performance measurement system examples.

Predictive Performance Measurement System Examples: A Traditional OEE Percentage Report-out

OEE or overall equipment effectiveness equates to equipment availability times performance times quality. However, this technicality is really not important for this first illustration of several predictive performance measurement system examples.

Management monthly received the following report-out as part of their performance measurement system (used with permission):

 

Performance Measurement System Traditional Report Example

Performance Measurement System Traditional Report Example

 

The question is what action or non action would be taken from this table of numbers. It is difficult to say what conclusion any person might make. However, let’s examine this data using a 30,000-foot-level predictive performance reporting methodology for attribute data. The benefit of a 30,000-foot-level approach is this methodology can:

  • Become a predictive performance decision making system
  • Examine the response from a process and improvement point of view.

Predictive Performance Measurement System Examples: A 30,000-foot-level OEE Percentage Report-out

A 30,000-foot-level report-out for Line 2, which would utilize historical data as part of a performance measurement system, is:

Performance Measurement System 30,000-foot-level Report for Line 2

Performance Measurement System 30,000-foot-level Report for Line 2

 

This 30,000-foot-level reporting is not the same as traditional control charting, which is to identify when unusual events occur in a manufacturing process so that operations can be halted to resolve the problem. This 30,000-foot-level reporting provides a high-level view of how the process is performing, where a prediction statement can be made if the process is stable. Points that should be noted with this form of performance measurement system reporting are:

  • The upper and lower limits shown in the chart (UCL and LCL) are statistically determined from the variability in the process. That is, not the response that is desired from the process. When points are outside these bounds the process is said to be non-stable; hence, it is not predictable. Week 7 had an unusual low point; hence, the overall process is said to be out of control or not stable.
  • When a process is not stable, an investigation should be undertaken for understanding and resolving the lack of stability. It should be highlighted that this type of action or consideration would virtually never be uncovered when one examines only the last weekly numerical response from a process.
  • One should also examine the control limits range; i.e., 50% – 89%. This is a very large range. Is this reasonable for this type of process response? Again, this consideration would not be uncovered with a simple table of numbers in a performance measurement system.

Let’s now examine a 30,000-foot-level report-out for Line 1:

 

Performance Measurement System 30,000-foot-level Report for Line 1 – No Staging

Performance Measurement System 30,000-foot-level Report for Line 1 – No Staging

 

Again, from a plot of this manufacturing line’s data, one would conclude if they were to examine all the data that this process was unstable; however, since week 26 the process appears to be more consist. Consider that an investigation revealed that something changed in week 26.

Because of this, we could then examine the data after this time period separately from before this weekly time period. The following chart stages the data for these two points in time, which would results in the following 30,000-foot-level chart. Data from the recent region of stability is used to formulate the prediction statement made at the bottom of the 30,000-foot-level report-out in this performance measurement system.

 

Predictive Performance Measurement System Examples: OEE 30,000-foot-level Report for Line 1 – Staging with Prediction Statement

Predictive Performance Measurement System Examples: OEE 30,000-foot-level Report for Line 1 – Staging with Prediction Statement

 

From this chart, we would now conclude that the process is stable with an OEE rate about 70.3%. From this chart, we understand that there will be week-to-week chance variation of the OEE response; however, the overall process is performing at a 70.3% OEE rate.

Predictive Performance Measurement System Examples: How to improve a 30,000-foot-level Predictive Process Statement

When a process is predictable, data from the recent region of stability can often be categorized to determine what areas might be given a concentrated effort to improve the overall process response. For this OEE situation, the following could be done to improve the overall company’s OEE response:

  • The above plot examined data from lines individually; however, there could have been an executive-level overall rate for all manufacturing lines combined. A sub-categorization could then be individual lines. From this performance measurement system assessment, one could evaluate whether specific lines should be given the most focus so that the overall OEE rate could be enhanced.
  • Since OEE equates to equipment availability times performance times quality, one can gain insight if there is an understanding which component of this equation causes the most negative impact to the overall OEE score. The manufacturing lines which were then targeted could conduct an analysis to determine which component, equipment availability, performance or quality, might be the best target for improving the overall OEE rate.

If a significant improvement change was made to improve a process’ overall response as part of this performance measurement system, the 30,000-foot-level individuals chart would transition to an enhanced level of performance.

Predictive Performance Measurement System Examples: Enhancement to Table of Numbers, Stoplight Scorecard, and Time-series Metric Reporting

Organizations benefit when the 30,000-foot-level predictive performance decision making process described above is applied to all forms of metric report-outs. Examples of reporting applications are:

Organizations benefit when they:

Businesses benefit when the described methodologies are applied to key performance measures throughout an organization. Business performance measurement applications include supply chain performance measurement and the monitoring of transactional processes.

Other dashboard conversions to predictive performance reporting illustrations are available in the article Transitioning Traditional Dashboards to Predictive 30,000-foot-level Metric Reporting Examples.

How to create Predictive Performance Measurement System Metrics

Additional information about the creation of 30,000-foot-level predictive performance decision making can be found:

Hopefully these predictive performance measurement system examples have been beneficial. Contact me directly to discuss the application of the described techniques to specific metrics or any overall performance management systems business application:

Looking forward to your comments.  Also, if you found this blog to be beneficial, it would be great if you tell others through this page’s social media links; i.e., Google+, etc.

Leave a Reply