Improving Overall Equipment Effectiveness (OEE) with PathWave Manufacturing Analytics

2021-07-29  |  8 min read 

With Industry 4.0 transformation, more analytics are performed on data as compared to previously where data is stored mainly for traceability and archive purposes.  More importantly, is how these data are being analyzed to improve product quality and increase production efficiency.  One can collect all kinds of data, but if these data are not analyzed properly, there will be no insight to drive improvement.

Overall Equipment Effectiveness (OEE in short) is a typical measure in manufacturing environment, which scores how effective the equipment is running.  The higher the OEE score, the more efficient the equipment is.  So how should one improve when an equipment’s OEE score is low?  Is it by testing more boards?  Is it by increasing shifts duration?  Or by providing more training to operator?

To answer that, let’s take a deeper look into OEE and how PathWave Manufacturing Analytics (PMA) provides insights to user for OEE improvement. OEE comprises of 3 components, namely Availability, Performance and Quality.

OEE = Availability x Performance x Quality

From the above formula, one can easily tell that each of these 3 components has direct impact to the OEE score.  Each component measures different aspects of the equipment on the production floor, which will be discussed in the following sections.   

Figure 1: OEE Dashboard with site summary view and drilldown equipment details

Figure 1 shows OEE dashboard in PMA.  It provides both high level Site OEE Summary as well as OEE score of each equipment; drilled down to Availability, Performance and Quality matrix.

In the industry, achieving OEE score of 85% is considered world-class, typically most will score an average between 60% to 80%.  Any score lower than 60% will be considered bad performance.


Availability Score (%) = (Utilization Time / Availability Time) x 100%

Availability Time is the amount of time scheduled for the equipment to operate.  This is typically link to the Shifts schedule of the production.  Operators reporting for each shift, there will be meal breaks and scheduled downtime for equipment maintenance. 

Utilization Time is the amount of time that equipment is utilized for actual production testing.

Ideally, if a shift is scheduled for 8 hours and the equipment is utilized fully for testing without any unplanned breaks, idle time, unplanned downtime, then the Availability score will be 100%.  Of course, this is easier said than done. 

Low Availability Score may be due to unexpected/unplanned downtime.  Meaning the equipment is unable to operate due to failures.  Frequent timely maintenance of equipment will definitely help to improve system uptime.

In PathWave Manufacturing Analytics (PMA), users can set up the Shifts & Schedule as well as the Planned Downtime.  It also includes a setting where user can set a duration for “Operator Handling Time”.  So that any duration between current board test and the next board test, coming below this Operator Handling Time setting, this duration will be considered as Utilization Time.  This helps to reflect more accurate Availability score in production.    

Figure 2: Downtime Monitoring Dashboard

In addition, with proper setup between the equipment and PMA, machine status of the equipment can be monitored, analyzed and presented in Downtime Monitoring dashboard (as shown in Figure 2).  It provides breakdown of Non-Utilization Time into Idle, Fixture Change, Equipment Down, etc. 

With these insights, user can take appropriate actions like adjusting SMT beat rate, investigation on long fixture change-over duration or increasing maintenance frequency to improve system uptime.


Performance Score (%) = [(Ideal Cycle Time x Total Count) / Utilization Time] x 100%

Ideal Cycle Time meaning under ideal condition, how long does it take to produce a unit. 

For example: it takes 30 seconds (including operator handling time) to test a board.  Hence, in an hour, the equipment would have tested 120 boards.  Low Performance Score may be due to untrained operator taking more time to handle the board, or may be due to unstable tests which caused more retest and resulting in increased test time, etc.

Figure 3: Digital Quality module - Test Result

In PMA, user can review the Cpk analysis for every test in a particular project (as shown in Figure 3).  Allowing user to easily sort and focus on those tests with worst Cpk values.  The scatter plot which displays the measurement trend provide valuable insight which direct user to the next right course of action, to improve the stability of tests.  In the above scatter plot, its portraits a bimodal measurement (2 distributions).  This is usually the case whereby dual vendor sources of component are being used.  Prompting user to investigate further with the component sources. 


Quality (%) = (Number of Passing Count / Total Count) x 100%

This is a measure of good (passing) boards produced.  If all boards are tested with zero defect, then the Quality score is 100%.  However, there bounds to be defects in manufacturing even if the percentage is very low.  And the ability to analyze and take appropriate corrective actions means a step closer to zero defects, or at least maintain high Quality score.

Figure 4: Top 5 Worst Test Names by Project

In PMA (as shown in Figure 4), user can review the Top 5 Worst Test Name for each project to understand the top failure pareto.  And if multiple equipment and fixture, or panel boards are involved, PMA allows easy side-by-side comparison between equipment/fixture/panels.  If failures occurred only on a particular equipment/fixture/cavity, then it is most likely a hardware issue which technicians/vendors can be called in to investigate.  If failures occurred across the multiple equipment/fixture/cavities, then it is most likely a component part issue or upstream process issue which further investigation can be conducted.

Figure 5: Compare by Equipment

From (Figure 5) compare by equipment charts, user can easily observe that failures occurred on one particular equipment.  In this scenario, it is most likely a hardware issue which user can further troubleshoot with equipment vendor.

With this, I hope you get to know how PMA can help you on your next OEE improvement effort.

Do visit for more information on PathWave Manufacturing Analytics.

That’s all for now.  Stay safe.