Tag: OEE Analysis

SPC for OEE

Some of our readers have expressed an interest in the application of Statistical Analysis Tools for OEE.  We have also reviewed various texts and articles that have expressed opposing views on the application of Statistical Tools with OEE data.

Our simple answer is this:

  •  At best, statistical tools should only be used on unique or individual processes.   Comparisons may be made (with caution) between “like” processes, however, even these types of comparisons require a thorough, in-depth, understanding of product mix, customer demand, and potentially unique process considerations.
  • At worst, it may be worthwhile to use statistical analysis tools for the individual OEE factors:  Availability, Performance, and Quality.

Negative trends may have Positive returns

While certain top level improvements can be implemented across the company, such as quick die change or other tool change strategies, their very integration may (and should) result in changes to the existing or current operating strategy.  A quick tool change strategy may provide for substantially reduced set up times and, as a result, operations may schedule shorter and more frequent runs.  In turn, the net change to OEE may be negligible or even lower than before the new change strategy was implemented.

A drop or decline in OEE does not necessarily translate to negative financial performance.  From a cash flow perspective, the savings may be realized through reduced raw material purchases, reduced inventories, and subsequently lower carrying costs that may more than offset any potential decrease in OEE.  As we have discussed in previous posts, OEE should not be regarded as a stand alone metric.  It is important to understand the financial impact of each of the OEE factors to your bottom line.  We’re in business to make money and Cash is King.

Scope of Analysis – Keep it Simple

As the scope of the OEE analysis increases from shift, to daily, to weekly, to monthly summaries, the variables that affect the end result are compounded accordingly.  Extending statistical techniques to OEE data across multiple departments or even company wide introduces even more sources of variation that make statistical modeling unrealistic.

While the application of statistics may sound appealing and “neat”, it is even more important to be able to understand the underlying factors that affect or influence the final result in order to implement effective countermeasures or action plans to make improvements or simply to eliminate the source of concern.

Regardless of the final OEE index reported, someone will ask the question, “So, what happened to our OEE?”  Whether an increase or decrease, both will require an answer to explain the variance.  If there was a significant increase, someone will want to know why and where.  This improvement, of course, will have to be replicated on other like machines or processes.  If there was a significant decrease, someone will want to know why and where so the cause of poor or reduced performance can be identified and corrected.

Ironically, no matter what the result, you will have to be prepared to supply individual process OEE data.  So why not just review the primitive data and respond to the results in real time?  While we recommend this practice, there may be certain trends relative to the individual factors that can be statistically evaluated in the broader sense (Quality – PPM, Labour – Efficiencies).

Statistics on a larger scale

We would suggest and recommend using statistical techniques on the individual Availability, Performance, and Quality factors of OEE.  For example, many companies track labour efficiencies relative to performance while others measure defects per million pieces relative to Quality.

While most people readily associate statistics with quality processes, many operations managers are applying statistical analysis techniques to a variety of metrics such as run time performance, performance to schedule, downtime, and setup times.  Maintenance managers are also analysing equipment availability for Mean Time to Repair, Meantime Time Between Failures, and equipment life cycle performance criteria.

As one example, we have conducted and encouraged our clients to consider statistical analysis of production data.  Tracking the standard deviation of daily production over time can reveal some very interesting trends.  These results will also correlate with the OEE factors.  Where the standard deviation is low, the increase in production is reflected in other metrics as well including financial performance.

A final note

Lastly, OEE is a tool that should be used to drive improvements.  As such, the goal or target is forever changing whether in small or large increments.  Another SPC solution for OEE that may be a little easier to understand and execute is as follows:

  1. System – Define and establish an effective system for collecting, analyzing, and reporting OEE data – preferably in real time at the source.
  2. Process – Understand and establish where and how OEE data will actually be collected in your processes and how it will be used to make improvements.
  3. Control– Establish effective methods to control both systems and processes to assure OEE is and remains a truly integrated metric for your operations.

We strongly recommend and support “at the source” thinking strategy.  Quite simply, we prefer points of control that are as close to their source as possible whether it be data, measurement, or product related.  Quality “at the source” (at the machine in real time) is much easier to manage than final inspection on the dock (hours or even days later).  Similarly, OEE managed in real time, at the process or machine, will serve the people and the company with greater control.

We welcome your feedback.  Please leave a comment or send us an e-mail (leanexecution@gmail.com) with your questions, suggestions, or comments.

Until Next Time – STAY lean!

We respect your privacy – we will not distribute, sell, or otherwise provide your contact or other personal information to any outside or third party vendors.

Advertisements

OEE Integration – Where do We Measure OEE? – Part I

OEE Integration Part IX – Where do we measure OEE?

Our recent posts have included numerous examples to calculate OEE correctly. We also discussed integration of OEE as an effective metric for managing your processes and ultimately how to analyze and use the data to improve your profitability.  We spent little time discussing where this measurement should occur.  OEE can be measured for both manual and automated lines as well as any stand alone operation.

The OEE factors (Performance, Availability, and Quality) are process output results.  The expectation, of course, is to manage the inputs to the process to assure the optimal result is achieved.  Availability, Performance, and Quality can be measured in real-time during production. However, the results should be subject to a due diligence review when production is complete.

At a minimum, it makes sense to measure OEE at the end (output) of the line or process but this is not always ideal.  The complexity of OEE measurement occurs when single or multiple sub-cells are constrained by an upstream or downstream operation or bottleneck operation.  The flow, rate, or pace of a process is always  restricted or limited by a  sequence / process constraint or bottleneck operation.  Just as a chain is only as strong as its weakest link, so too is the line speed limited by the bottleneck operation.

We contend that the “Control-Response” loop for any process must enable immediate and effective corrective action based on the measured data and observations.  Measuring OEE in real-time at the bottleneck process makes it an ideal “Trigger Point” metric or “Control-Response” metric for managing the overall process even in “isolation” at the bottleneck operation.  Any variations at the bottleneck correlate directly to upstream and downstream process performance.

A disruption to production flow may occur due to a stock-out condition or when a customer or supplier operation is down.  While these situations affect or impact the OEE Availability factor, external factors are beyond the scope of the immediate process.

Real-time OEE requires that these events and others, such as product disposition, are reported in real-time as well.  External events are more difficult to capture in real-time and by automated systems in particular.  Operator interfaces must accommodate reporting of these events as they occur.

Reporting PITFALL – After-the-Fact events

If a quality defect is discovered several days after reporting production and all parts are placed on hold for sorting or rework, the QUALITY Factor for that run should be changed to ZERO.  In turn, the net OEE for that run will also be ZERO.  If the system is not changed, the integrity of the data is lost.  This also exemplifies that real-time data can be deceiving if proper controls are not in place.

“Where do we measure?” is followed by “When do we measure?” The short list of examples provided here are likely events that are far and few between.  If this is a daily occurrence, consider adopting the banking policy of, “adjustments to your account will be reflected on the following business day”.  Your process / system is in need of a rapid fix.

OEE is one of the few vital signs or key performance metrics for your manufacturing operation.  As such, measure where you will reap the greatest benefit and focus your attention on the process or operation accordingly.  OEE is as much a diagnostic tool as it is a monitoring tool.

Until Next Time – STAY lean!

Vergence Analytics
Versalytics

Availability and OEE

What is Availability?

In its simplest form, availability measures the uptime of a machine or process against the planned production time.  As one of the factors of Overall Equipment Efficiency (OEE), Availability is expressed as a percentage.  The uptime is calculated by taking the difference between the planned production time and total duration of the downtime events that occurred during the planned production period.

We specifically address the “Availability” factor in this post for the simple reason that the definition of availability is likely to be one of the most debated and hotly contested topics of your OEE implementation strategy.  The reason for this, in many cases, is the lack of clarity in some of the most basic terminology.  The purpose of this discussion is to present some topics for consideration that will allow you to arrive at a clear definition that can perhaps be formed into a standard policy statement.

We will also demonstrate that it is possible to calculate the downtime by simply knowing the cycle time or process rate, the quantity of parts produced, and the planned production time.  We recommend using this technique to validate or reconcile the actual documented downtime.  We would argue that the first and foremost purpose of any machine monitoring or downtime event measurement system is to determine the “WHY and WHAT” of the downtime events and secondly to record the “When and How Long”.

You will learn that monitoring your processes to determine causes and duration of downtime events  is key to developing effective action plans to improve availability.  The objective of any machine automation, sensor strategy, or data collection and analysis is to determine methods and actions that will improve the availability of the equipment through permanent corrective actions, implementing more effective trouble shooting strategies (sensor technologies), improved core process controls, or more effective preventive maintenance.

Define the purpose of OEE

While it looks like we’re taking a step back from the topic of discussion, bear with us for just a paragraph or two.  A clear statement of purpose is the best place to start before executing your OEE implementation strategy:

To identify opportunities to improve the effectiveness of the company’s assets.

You will quickly realize that, when attempting to define the measurement criteria for the OEE factors, in particular Availability, your team may present rationale to exclude certain elements from the measurement process.  These rationalizations are typically predicated on existing policy or perceived constraints that simply cannot be changed.  People or teams do not want to be penalized for items that are “out of their control” or bound by current policy.  Continuous improvement is impeded by attempts to rationalize poor performance.

We understand that some of these “exclusions” present a greater challenge, however, we do not agree with the premise that they cannot be improved.  Again, it is a matter of “purpose”.  Limiting the scope of measurement will limit the scope of improvement.  Now it’s time to explore what could be the foundation for a sound definition of availability.

Availability Considerations

It may seem reasonable to assume that, at a minimum, the only planned down time events that should be excluded from the availability factor are  planned preventive maintenance activities, mandatory break periods, and scheduled “down” time due to lack of work.  We would argue and agree that the only justification for an idle machine is “Lack of Work”.

What would be the reason to settle for anything less?  If Preventive Maintenance is critical to sustaining the performance of your process, doesn’t it make sense to consider it in the measurement process?  The rationale that typically follows is that Preventive Maintenance must be done and it’s really out of our control – it is a planned event.  We would argue that the time to complete Preventive Maintenance can be improved.

Is it possible that the Mean Time Before Failure or Required Maintenance can be extended?  Is it possible to improve materials, components, or lubricants that could extend the process up time?  Is it possible to improve the time it actually takes to perform the required maintenance?  If so, what is the measure that will be used to show that additional capacity is available for production.

If set up times for die changes or tool changes can be improved from hours to minutes, could the same effort and devotion to improve Preventive Maintenance techniques yield similar results?  We think so.

One example is the use of synthetic oils and lubricants that have been proven to significantly extend the life of tools and components and also reduces the number changes required over the service life of the machine.  Quick change features that can assist with easy and ready access to service points on tooling and machines can also be implemented to reduce preventive maintenance times.

The other exclusion that is often argued is break times.  Labour laws require you to provide break times for your employees.  However, since automated processes are not subject to “Labour Laws”, the “mandatory break times” do not apply.  We would argue that methods should be pursued to reduce the need for human intervention and look for ways to keep the machine running.  Is it possible to automate some of the current processes or rotate people to keep the machine running?

Aside from this more obvious example, consider other organizational policies that may impact how your organization runs:

  1. Shift start-up meetings
  2. Employee Communication Meetings
  3. End of Shift clean up periods
  4. Quality first off approval process
  5. Shift first off versus Run first off
  6. Weld Tip changes – PM or Process Driven

 What is the purpose of the shift start-up meeting?  What is the purpose of the monthly employee communication meeting?  Could this information be conveyed in a different form?  What length of time is really required to convey the message to be shared?  Is the duration of the meeting actually measured or do you resort to the standard time allotted?

Clean up periods at the end of the shift  are also a common practice in many plants.  What is being cleaned up?  Why?  Is it possible to maintain an orderly workplace during the shift – clean up as it happens in real-time?  Again, do you record the actual clean up time or do you just enter the default clean up time allotted?

How much time is lost to verify the integrity of the product before allowing production to commence?  What process parameters or factors would jeopardize the quality of the product being produced?  No one wants to make scrap or substandard components, however, the challenge remains to determine what factors influence the level of quality.  If it is possible to determine what factors are critical to success in advance, then perhaps the quality verification process becomes a concurrent event.

Measuring Downtime.

 There are other factors that can impact availability including, but certainly not limited to, personnel (illness, inclement weather), material availability, other linked processes (feeder / customer), material changes, tool changes, quality concerns, and unexpected process, equipment, or machine faults.

It is possible to use manual or automated systems to collect various machine or process codes to record or document the duration and type of downtime event.  We recommend and support the use of automated data collection systems, however, they should be implemented in moderation.  One of the primary impediments to success is overwhelming volumes of data that no one has the time to analyze.

The Goal = 100% Up Time = ZERO Down Time = Zero Lost Time = Zero Defects = 100% Availability

The goal is to use the data and tools available to either permanently resolve the problem by implementing an effective corrective action or to assist the trouble shooting process by identifying the failure mode and to minimize the duration of the downtime event.

We have witnessed data collection strategies where an incredible number of sensors were installed to “catch” problems as they occur.  The reality was the sensors themselves became the greater cause of downtime due to wear or premature failure due to improper sensor selection for the application.  Be careful and choose wisely.

When used correctly, automation can be a very effective tool to capture downtime events and maintain the integrity of the overall measurement process.  With the right tools, trouble shooting your process will minimize the duration of the down time event.  Monitoring the frequency of these events will also allow you to focus your attention on real opportunities and circumvent nuisance faults.

The objective of collecting the “downtime event” history is to determine what opportunities are available to improve uptime.

Duration versus Frequency

The frequency of a downtime event is often overlooked as most of the attention is devoted to high duration downtime events.  Some sources suggest that short duration downtime events (perhaps as little as 30 seconds) are not worth measuring.  These undocumented losses are reflected, or more accurately hidden, by a corresponding reduction in the performance factor.

Be careful when setting what appears to be simple policy to document downtime.  A 20 second downtime event that occurs 4 times per hour could quickly turn into 10 minutes a shift, 30 minutes a day, 2.5 hours a week, 125 hours a year.  Rather than recording every event in detail, we recommend implementing a simple “tick” sheet to gain an appreciation for the frequency of failures.  Any repetitive events can be studies and reviewed for corrective action.

Verify the Downtime

One of the advantages of OEE is that it is possible to reconcile the total time – OEE should never be greater than 100%.  Of course this statement requires that the standard cycle time is correct and the total quantity of parts produced is accurate.  So, although all of the downtime events may not be recorded, it is very easy to determine how much downtime occurred.  This will help to determine how effectively downtime data is being recorded.

A perfect example to demonstrate this comes from the metal stamping industry.  Progressive dies are used to produce steel parts from coil steel.  The presses typically run at a fixed “predetermined” optimum run rate.  Depending on the type of part and press, progressive dies are capable running at speeds from as low as 10 strokes per minute up to speeds over 300 strokes per minute.

For ease of calculation, assume we have a press that was scheduled to run a part over an 8 hour shift having two 10 minute breaks.  The standard shift hours are 6:45 am – 3:15 pm and 3:30 pm – 12:00 am.  The company provides a 30 minute unpaid meal break after 4 hours of work.  The optimum press speed to run the part is 20 strokes per minute (spm).  If a total of 6200 parts were made – how much downtime was incurred at the press?

To determine the press time required (also known as earned time), we simply divide the quantity of parts produced by the press rate as follows:

Machine Uptime:  6200 / 20 = 310 minutes

Our planned production time was 8 hours or 480 minutes.  Assuming that company policy excludes break times, the net available time to run the press is 480 – (2 x 10) = 460 minutes.

Calculated downtime = Available – Earned = 460 – 310 =150 minutes

Availability = Earned Time / Net Available Time = 310 / 460 = 67.39%

We can see from the above example that it easy to determine what the downtime should have been and, in turn, we could calculate the availability factor.  This calculation is based on the assumption that the machine is running at the stated rate.

The Availability TWIST (1):

Knowing that press and die protection technologies exist to allow presses to run in full automatic mode, the two break periods from our example above do not apply to the equipment, unless company policy states that all machines or processes must cease operations during break periods.

Assuming that this is not the case, the press is available for the entire shift of 480 minutes.  Therefore, the availability calculations from above would be:

Calculated downtime = Available – Earned = 480 – 310 =170 minutes

Availability = Earned Time / Net Available Time = 310 / 480 = 64.58%

The Availability TWIST (2):

Just to expand on this concept just a little further.  We also indicated that the company provided an unpaid lunch period of 30 minutes.  Since meal breaks don’t apply to presses, the reality is that the press was also available to run during this period of time.  The recalculated downtime and availability are:

Calculated downtime = Available – Earned = (480 + 30) – 310 =200 minutes

Availability = Earned Time / Net Available Time = 310 / 510 = 60.78%

The Availability TWIST (3):

Finally, one last twist (we could go on).  We deliberately indicated that there was a 15 minute break between shifts.  Again, is there a reason for this?  Does the machine have to stop?  Why?

Availability – NEXT Steps

As you begin to look at your operations and policies, start by asking WHY do we do this or that?  The example provided above indicates that a significant delta can exist in availability (close to 7%) although the number of parts produced has not changed.  The differing results are related to policy, operating standard, or both.

If the performance (cycle time or production rate) and total quantity of parts produced data have integrity, the availability factor can be reconciled to determine the integrity of the downtime “data collection” system.  From this example it should also be clear that the task of the data collection system is to capture the downtime history as accurately as possible to determine the opportunities to improve availability NOT just to determine how much downtime occurred.

This example also demonstrates why effective problem solving skills are critical to the success of your lean implementation strategy and is also one of the reasons why programs such as six sigma and lean have become integrated as parallel components of many lean execution strategies.

The Goal:  100% uptime / Zero downtime / Zero lost time /100% availability

Regardless of the measurement baseline used, be consistent.  Exclusions are not the issue, it is a matter of understanding what is involved in the measurement process.  For example, maintenance activities performed during break periods may be a good management practice to improve labour efficiencies, however, the fact that the work was performed during a break period should not exclude it from the “downtime” event history.  We would argue that all activities requiring “equipment time” or “process time” should be recorded.

FREE Downloads

We are currently offering our Excel OEE Spreadsheet Templates and example files at no charge.  You can download our files from the ORANGE BOX on the sidebar titled “FREE DOWNLOADS” or click on the FREE Downloads Page.  These files can be used as is and can be easily modified to suit many different manufacturing processes.  There are no hidden files, formulas, or macros and no obligations for the services provided here.

Please forward your questions, comments, or suggestions to LeanExecution@gmail.com.  To request our services for a specific project, please send your inquiries to Vergence.Consulting@gmail.com.

We welcome your feedback and thank you for visiting.

Until Next Time – STAY Lean!

"Click"

Variance, Waste, and OEE

What gets managed MUST be measured – Including VARIANCE.

It is easy to get excited about the many opportunities that a well implemented LEAN Strategy can bring to your organization.  Even more exciting are the results.

Achieving improvement objectives implies that some form of measurement process exists – the proof.  A clear link should be established to the metric you choose and the activity being managed to support the ongoing improvement initiatives.

Measure with Meaning

Why are you “collecting” OEE data?  While OEE can and should be used to measure the effectiveness of your manufacturing operations, OEE on its own does not present a complete solution.  It is true that OEE presents a single metric that serves as an indicator of performance, however, it does not provide any insight with respect to VARIANCES that are or may be present in the system.

We have encountered numerous operations where OEE data can be very misleading.  OEE data can be calculated using various measurement categories:  by machine, part number, shift, employee, supervisor, department, day, month, and so on.

VARIANCE:  the leading cause of waste!

Quality professionals are more than familiar with variance.  Statistically capable processes are every quality managers dream.  Unfortunately, very little attention or focus is applied to variances experienced on the production side of the business.

Some may be reading this and wonder where this is going.  The answer is simple, rates of production are subject to variance.  Quite simply, if you review the individual OEE results of any machine for each run over an extended period of time, you will notice that the number is not a constant.  The performance, availability, and quality factors are all different from one run to the next.  One run may experience more downtime than another, a sluggish machine may result in reduced in performance, or material problems may be giving rise to increased quality failures (scrap).

So, while the OEE trend may show improvement over time, it is clear that variances are present in the process.  Quality professionals readily understand the link between process variation and product quality.  Similarly, variation in process rates and equipment reliability factors affect the OEE for a given machine.

We recommend performing a statistical analysis of the raw data for each factor that comprises OEE (Availability, Performance, and Quality) for individual processes.  Analysis of OEE itself requires an understanding of the underlying factors.  It is impractical to consider the application of ANOVA to OEE itself as the goal is to continually improve.

How much easier would it be if you could schedule a machine to run parts and know that you will get them when you needed them?  You can’t skip the process deep dive.  You need to understand how each process affects the overall top-level OEE index that is performance so you can develop and implement specific improvement actions.

The best demonstration we have seen that illustrates how process variation impacts your operation is presented through a “process simulation” developed from Eli Goldratt’s book, The Goal.  We will share this simulation in a separate post.  Experiencing the effect of process variation is much more meaningful and memorable than a spreadsheet full of numbers.

Conflict Management and OEE

In some environments we have encountered, the interpretation of LEAN strategy at the shop floor level is to set minimum OEE performance objectives with punitive consequences.  This type of strategy is certainly in conflict with any Lean initiative.  The lean objective is to learn as much as possible from the process and to identify opportunities for continual improvement.

Management by intimidation is becoming more of a rarity, however, we have found that they also give rise to the OEE genius.  If performance is measured daily, the OEE genius will make sure a high performing job is part of the mix to improve the “overall” result.  This is akin to taking an easy course of study to “pull up” your overall average.

It is clear from this example, that you will miss opportunities to improve your operation if the culture is tainted by conflicting performance objectives.  The objective is to reveal sources of variation to eliminate waste and variation in your process, not find better ways to hide it.

Variance in daily output rates are normal.  How much are you willing to accept?  Do you know what normal is?  Understanding process variance and OEE as complementary metrics will surely help to identify more opportunities for improvement.

FREE Downloads

We are currently offering our Excel OEE Spreadsheet Templates and example files at no charge.  You can download our files from the ORANGE BOX on the sidebar titled “FREE DOWNLOADS” or click on the FREE Downloads Page.  These files can be used as is and can be easily modified to suit many different manufacturing processes.  There are no hidden files, formulas, or macros and no obligations for the services provided here.

Please forward your questions, comments, or suggestions to LeanExecution@gmail.com.  To request our services for a specific project, please send your inquiries to Vergence.Consulting@gmail.com.

We welcome your feedback and thank you for visiting.

Until Next Time – STAY Lean!

"Click"