Modern Production Metrics – OEE

  

Overall Equipment Effectivity:

OEE, Overall Equipment Effectivity, is the most popular KPI, Key Performance Indicator, in manufacturing productivity philosophies such as WCM, World Class Manufacturing, TPM, Total Productive Maintenance etc.

OEE is a measure of a machine’s effectiveness based on six general loss classifications, i.e. Breakdowns, Changeovers, Short Stops, Slow Running(Speed Losses), Yield Losses and Start-up Losses. These losses are grouped under the general headings; Availability Losses, Performance Losses and Quality Losses as described below.

Put simply, OEE is a measure of the amount of time a machine is actually producing good product compared to the time it should be producing good product (the total planned production time for that machine). Any difference between these two measures of time is considered a loss.

Figure 1 shows an OEE Level 1 Pie Chart Report which divides the loss into three groups, Availability, Performance and Quality.

 

Fig. 1 OEE Level 1 Pie Chart Report (click image to see full size).

The Six Losses are listed and categorized as follows:   

  • Availability Losses.
    Availability Loss is a general term for any loss which causes a machine to be unavailable to produce good products.

    1. Breakdowns:
      When a machine has broken down it is unavailable for production. In some OEE systems a machine is only considered to be ‘broken down’ if a technician is required to restart the machine. However this would require manual classification of each loss and in automatic OEE Production Monitoring Systems manual classification is considered cumbersome. Consequently the compromise is generally made that if the downtime duration is longer than, say, 5minutes the downtime is automatically classified as a Breakdown, otherwise the downtime is classified as a Short Stop.
       
    2. Changeovers:
      When a machine has broken down it is unavailable for production. In some OEE systems a machine is only considered to be ‘broken down’ if a technician is required to restart the machine. However this would require manual classification of each loss and in automatic OEE Production Monitoring Systems manual classification is considered cumbersome. Consequently the compromise is generally made that if the downtime duration is longer than, say, 5minutes the downtime is automatically classified as a Breakdown, otherwise the downtime is classified as a Short Stop. 

 

  • Performance Loss.
    Performance Loss is a general term for a loss occurring during production which reduces the performance of the machine. Performance Loss is sometimes also referred to as Speed Loss.

    1. Short Stops:
      When a machine is in production and it stops for a short period of time for a minor fault that the operator can correct in a few seconds, this is called a Short Stop. If an operator is operating more than one machine these Short Stops are highly significant. It may only take a few seconds to fix the machine, but if the operator is busy with another machine or taking a break, then the machine will be stopped for a significant amount of time. If these minor faults occur frequently, then over the course of a day the amount of time lost can accumulate quite rapidly. This is especially true where an operator is running a large number of machines as it will not be possible to keep them all running if there are a lot of minor faults happening simultaneously.
       
    2. Speed Loss(Slow Running):
      This loss is not usually apparent by simply looking at a machine. Generally when a machine is running and producing good parts, it may appear that all is well. However there may still be a loss occurring if the machine is operating below its designed speed. That is, if the machine is designed to produce 1000 parts per hour but for some reason is actually only producing 750 units per hour then it is only running at 75% of its capability and there is a 25% loss due to Slow Running.

 

  •  Quality Losses.
    Quality Loss is the general term for the time lost producing bad or reject parts.

    1. Yield Loss:
      When a machine produces a defect not only is the material used in producing the defect lost or in need of rework but the time and other resources used producing the piece are also wasted. A machine with ten defects per hundred is in effect only achieving a yield of ninety per cent of its capability.
       
    2. Start-up Loss:
      If a machine needs to be set up by doing some trial production then the material used is wasted. For example setting up a machine at the start of a shift could involve producing one or two test pieces and then making adjustments until the set up is perfect. The material lost and the time spent producing it are both wasted and again this is a problem. 

 

Key Implications:

  • Classifying the losses in terms of Availability, Performance and Quality helps to identify the nature of the most significant losses effecting production.
     
  • From knowing the nature of the most significant losses it becomes easier to identify the root causes and to implement strategies to eliminate the losses.
     
  • It’s only when you multiply the 3 elements, Availability, Performance Rate and Quality Rate together that you see the compound effect.
     
  • This compound effect often highlights a surprisingly aggressive erosion of OEE! (Remember that two fractions multiplied together will always result in a smaller fraction!)


Calculating OEE:

OEE may be expressed as a time value or as a percentage (rate). Below it is shown how to calculate the OEE value as either a time or a percentage.

Expressing OEE as a Time Value:

The figure below shows the losses occurring on a machine over a given time frame.

 

Fig. 2 OEE Time (click image to see full size).

    • Time Frame
      The period over which the OEE is to be calculated.
       
    • Planned Operating Time
      The amount of time for which it is planned that the machine should be producing good parts. It is the Time Frame less any planned downtime such as preventative maintenance or operator training etc.
       
    • Availability Loss
      The total amount of time for which the machine has been broken down or on changeover. 
       
    • Actual Operating Time
      The Planned Operating Time less the Availability Loss.
       
    • Performance Loss
      The combination of Short Stops and Slow Running.
       
    • Net Operating Time
      The Actual Operating Time less the Performance Loss.
       
    • Quality Loss
      The amount of time lost producing bad or reject parts. It is calculated by multiplying the number of reject parts by the optimum (design) machine cycle time.
       
    • OEE Time (Fully Productive Time)
      The total amount of time that the machine was operating at its optimum or designed rate. It is the Planned Operating Time less the Availability Loss, less the Performance Loss and less the Quality Loss. The OEE Time is also known as the Fully Productive Time.

 

Expressing OEE as a Percentage:

Which in graphical format can be represented by figure 3 below.

 

Fig. 3 OEE Time as Percentage (click image to see full size).

    • Availability (Rate)
      The percentage of time the machine is actually available to produce good parts. In other words, the Actual Operating Time (Planned Operating Time – Availability Loss) compared to the Planned Operating Time.


       

    • Performance (Rate)
      The Net Operating Time (Actual Operating Time – Performance Loss) compared to the Actual Operating Time.


       

    • Quality (Rate)
      The OEE Time (Net Operating Time – Quality Loss) compared to the Net Operating Time. The Quality Rate is equivalent to the Yield.


       

Interpreting OEE Values:

As you will see from figure 2, Availability, Performance and Quality Losses eat into the Planned Operating Time for the machine. The greater the losses the less output from the machine. The aim therefore is to keep these losses to a minimum.

This is equivalent to saying that the Availability, Performance and Quality Rate must be kept as high as possible.

If, for example, the Availability (Rate) is 0%, this means that the machine was either broken down or on changeover for the whole time period. Whereas if the Availability is 100% it means there were no breakdowns or changeover losses during the time period.

The key value is the percentage OEE. This indicates how well the machine is performing. The higher the value the better the machine is performing. An OEE of 100% indicates that there were no losses during the time period and that the machine was running at its optimum rate.

The OEE value that can be achieved for a particular machine will depend on a wide variety of factors but in general companies tend to aim for an OEE value of between 70 – 90%.

The purpose of separating losses into different categories is to help focus on the reasons and effects of the different types of losses. The reasons for breakdown losses are often for totally different reasons then reject losses.

Availability Losses will generally be related to poor machine reliability, bad maintenance or overloading the machine causing failure.

Performance Losses can often relate to material problems requiring the machine speed to be reduced, poor design requiring regular operator intervention for minor stoppages, poor operation due to insufficient operator training.

Quality Losses can be related to faulty raw material, or machine problems.

 

OEE Loss Levels:

There are three levels in the hierarchical OEE model of grouping losses.

Level 1 Losses are Availability, Performance and Quality losses.

Figure 4 shows a Pie chart type OEE Level 1 Loss Report. Each slice represents the fraction of Total Loss which is attributed to each category.

  Fig. 4: OEE Level 1 Pie Chart Report (click image to see full size).

 

Level 2 Losses are simply Level 1 Losses sub-divided by the six major losses defined above, Breakdowns, Changeover’s, Short Stops, Slow Running, Yield Losses and Start-up Losses.

Figure 5 shows a Pie chart type OEE Level 2 Loss Report. Again each slice represents the fraction of Total Loss which is attributed to each category.

  Fig. 5: OEE Level 2 Pie Chart Report (click image to see full size).

 

Level 3 Losses are the individual reasons for each Level 2 Loss. An example of a Level 3 Loss might be “jam on station 1” where this is a specific reason for the machine to stop – causing a loss.

Figure 6 shows OEE Level 3 Pie Chart Loss Report. Here each slice represents the fraction of Total Loss which is attributed to each individual loss reason.


  Fig. 6: OEE Level 3 Pie Chart Report (click image to see full size).

 

Using Provideam to identify and prioritise improvement activity.

Provideam provides the tools to analyse your production data to identify the real reasons for loss.  Thus you can focus improvement activities on the eliminating these ‘real’ losses.  Thereby ensuring that you do not waste time eliminating ‘perceived’ losses which are not actually significant in the overall productivity equation.

If you have any queries or would like further information, please do not hesitate to contact [email protected]

—————————-
 

 

Using Automation IT to Reach Your Productivity Targets

A Case Study on the Deployment of a Provideam OEE/Downtime Monitoring Solution

 

In this paper Paul Mitchell, MD of DTL Systems provides a case study of a typical implementation of a Provideam Solution in a Medical Devices Company. Provideam is a leading edge Manufacturing Productivity Solution developed by DTL Systems. Provideam generates Management Metrics for Lean Manufacturing Companies and is the chosen productivity solution of some of the World’s leading Life Sciences and General Manufacturing Companies.

 

 

“The Provideam Solution has been a fundamental tool in helping us achieve our annual savings and targets.”

John Quirke, Process Excellence Manager for Millipore UK and Ireland.

 

 

In recent years OEE has become the de facto standard metric of machine and process performance. Companies who use this metric to support a Continuous Improvement, Lean Management Style consistently out perform their competitors. Originally OEE was calculated manually from handwritten downtime and yield logs. Today, due to the enabling advances in Automation IT (Information Technology), manufacturers are collecting productivity data directly from production machines in real-time.

 

The benefits of Provideam, an automatic data collection and analysis solution, include;

  • Low cost, reliable and most importantly impartial data collection
  • Powerful Cross-Referencing/Querying Engine
  • Potential to Store and Analyse Large volumes of Data
  • Automatic Email Delivery of Shift, Day, Week Reports etc.

 

 

Approach:

Provideam is designed to be a plant-wide solution and can comfortably integrate data from all areas of your plant; from moulding to assembly to packing. However at DTL we strongly advise our customers to pilot the technology in one key area before attempting a roll-out on a plant-wide basis. The pilot enables the customer to carefully manage the technological and managerial implications of the deployment and it should ensure that the application is deployed in a way that best meets the requirements of the end user (Production, Quality and Technical Departments).

 

The goals of the pilot project should be well defined and relatively easy to achieve. These goals may not encompass every long-term objective you have for automated production monitoring – but that’s OK, once your pilot is successful it will be easy to refine the system over time, especially once you’ve had an opportunity to understand the capability of the system – that’s the foundation of Continuous Improvement.

 

Even though your pilot project shouldn’t require a huge capital investment, it should provide live, useful data which can immediately be adopted into the management of your machinery. Following a successful pilot it is a relatively straight forward task to implement the system on other similar machines. Note also that the data generated by your pilot will also be extremely useful for ‘selling’ the concept of Lean Manufacturing/Automated OEE Monitoring to other area of the plant. Provideam is sold directly to end-users or through VARs but in all cases we strongly advise on an initial pilot implementation.

 

 

Case Study:

Client:

Millipore Cork Ireland is an excellent example of a company that has implemented a Provideam Manufacturing Productivity Solution in support of its adoption of Lean Manufacturing Management. The combination of Lean Management and Provideam has enabled Millipore Cork to dramatically increase its productivity.

 

Millipore Cork Ireland makes a wide variety of tools to support life science research and is best known for its Filter Products. Manufacturing processes in Ireland include Injection Molding, Membrane Production, Product Assembly and Packing.

 

Around five years ago Millipore Cork management, recognising the ever more competitive global market place, made a strategic decision to adopt Lean Manufacturing. Following this decision, one of the key product assembly areas was selected for a trial of the concepts. From the outset Millipore Cork took the approach of simultaneously deploying Lean Management structures and automating the generation of productivity data. DTL Systems was selected to implement Provideam because of our integration expertise; our extensive experience of data collection and our appreciation of the operational and data requirements of Production Supervisors and Operators.

 

Pilot Project:

The trial area consisted of 10 independent assembly machines. The machines were all quite similar in terms of the functions they performed. However the control systems were not standard. The newer machines had Allen-Bradley PLCs (SLCs) with PanelView MMIs, the older machines had a variety of PLCs of varying vintages. Cork had a simple data capture network in place before the implementation of Provideam and it was the obvious decision to ‘piggy-back’ off this system. The data capture system consisted of a central SLC PLC communicating with the newer SLC PLCs over Ethernet and with the older PLCs via remote I/O. The Provideam Server Application contains a Data Collection Service which leverages OPC to connect to the widest variety of devices and systems. Although the Provideam OPC client will connect to many brands of OPC server, the single server with plug-in driver architecture of Kepware’s OPC server suite KEPServerEX allows Provideam to seamlessly connect to over 100 types of PLCs and systems all through a single consistent server interface. For this pilot, Kepware OPC Server with drivers for Allen Bradley was selected to communicate with the Data Collection PLC. Any programming required to prepare the raw data for data collection was done by Millipore Cork Engineers and Millipore Cork Controls Engineering Sub-contractors. The majority of this programming was carried out in the Data Collection PLC, rather than the Control PLCs. This has made it easier to implement improvements as there is rarely a need to modify the Control PLCs.

 

The Provideam Data Collection PC sends the data to a database in the Provideam Server and from there it is made available to users via an Intranet Application which can be accessed anywhere in the plant.

 

 

 

 

Fig. 1: Data Collection Schema of Cork Pilot Implementation

 

Extension of Pilot Project:

As an extension to the Pilot Project a number of Filter Cutting Machines were added to the Provideam System. These were each controlled by a proprietary industrial PC which was not accessible to Millipore Cork. The only interface to the control PC was an OPC Server provided by the manufacturer. The challenge for Millipore Cork was to convert the raw data generated by this OPC Server into a profile which modelled the operational mode of each machine. The solution Millipore Cork came up with was to map the raw data from the machines into the existing Data Collection PLC via Kepware’s LinkMaster Application. In the Data Collection PLC the raw data was easily manipulated to model the operating condition of each machine. This avoided the requirement to modify any logic in the control PCs.

 

 

 

Typical Screenshots:

 

Fig. 2: Real-time Dashboard – Current Status

 

Fig. 2 shows a typical Provideam Dashboard. The Key Performance Indicators for two machines over two shifts are displayed. The lower section of the screen shows the real-time details of the selected machine. In the above example the green lamp indicates that the machine is running.

 

Fig. 3: Downtime Log

 

Fig. 3 shows the downtime log for a selected machine. The time and duration of each stop is shown. This is very useful and enables a technician to evaluate how the machine is performing and in what sequence downtimes are occurring.

 

Fig. 4: OEE Time

 

Fig. 4 shows a graphical representation of the OEE Time for each hour of a shift for a selected machine. In this example the machine appears to be running very well. This type of screen is very useful for production supervisors to see how production is going over the course of a shift.

 

Fig. 5: OEE Analysis

 

Fig. 5 shows a typical pie OEE Analysis for a selected machine. In this example we can see that the Speed Loss is the main productivity loss. This enables the Lean Team to focus in on the main causes of productivity loss.

 

The above screenshots represent a small sample of what is available in Provideam. All the data displayed can also be shown in tabular form.

 

In the Provideam Reporting module it is possible to create customised reports for a Month, a Week, or a Customised Period etc. A very useful feature of Provideam is that reports can be scheduled to be delivered to anyone by email as soon as the shift, month etc is completed. Therefore you can get your reports sent to you without ever having to log on to Provideam.

 

Phase 2:

The initial machines were relatively straight forward as there was no requirement to monitor changes in Part, Tool, Lot etc. For these machines it was only necessary to know the downtime reasons and the yield. These machines typically had 100 or so downtime reasons and 3 defect reasons. The next area to implement Provideam was the Moulding Area and this was a much more complex environment requiring Lots, Material, Tools, Parts, Cavities etc. to be monitored.

 

The simplest method of collecting data from the moulding machines was to deploy ProvDAQ Pushbutton stations, see Fig. 6. When the machine stopped the ProvDAQ lamps flashed, prompting the operator to select a downtime reason by pressing the appropriate button on the ProvDAQ.

 

Fig. 6: ProvDAQ – Pushbutton Data Acquisition Terminal

Pushbuttons are labelled with each appropriate downtime or changeover reasons.

 

That button selection stayed illuminated until the machine was restarted. When changing from one Lot to another the operator simply pressed a ‘ChangeOver’ button. This caused a new Lot record to be created in Provideam. The Lot details were updated, in Provideam, by the Production Supervisor once the machine was running. Although this may appear an unsophisticated solution it is highly effective and quick, and avoids the need for the operator to leave the machine to record downtime data.

 

Roll-out:

Following successful implementation in the assembly and moulding areas, Provideam was installed in each successive area as Lean Management was introduced to that area. Each area presented some engineering challenges in terms of how to acquire data from the machines and these were addressed by matching the sophistication of the equipment with a data collection structure to suit. For example, in areas where the operators control the equipment via SCADA, Cork Controls Engineers integrated Provideam Data Capture into the SCADA Interface.

 



Outcomes:

  1. Real-time Data – Provideam provides real-time data which allows Millipore Cork to react immediately if productivity starts to drop off. In manual reporting systems it can be hours if not days before it is realised that there has been a problem.
  2. Accurate Objective Data – Provideam data is generated by the actual control system and is therefore not subjective. Note our experience is that the majority of productivity losses are due to bad machine design, bad process organisation and poor technical support rather than inefficient operators.
  3. Consistent Data – In Lean Manufacturing, multi-disciplinary teams work together and constantly strive to improve performance. Having one source of productivity data ensures that the focus is on resolving productivity issues, rather than arguing about the data. It is also a great advantage to have the whole plant use the same productivity solution as this facilitates the use of a same metric system in all areas allowing you to compare different functional areas. It also offers plant-wide data via one system rather than having to rely on and compare data collection from multiple disparate systems.

 

Conclusion:

Provideam is used daily by, Production, Maintenance, Planning, and Process Engineering.

 

It enables the users to identify new areas for improvement activity while at the same time helping to lock-in productivity gains already achieved.

 

 

“The Provideam Solution has enabled us to increase Productivity by between 20 and 30% at the Millipore Cork Facility over the last 5 years. In addition, the constant monitoring provided by Provideam has ensured that these increases are maintained.”

John Quirke, Process Excellence Manager for Millipore UK and Ireland.

 

 

Recap:

The intention of this article is to demonstrate that reliable metrics for continuous improvement can be introduced very effectively by means of focused pilot projects.

The best pilot projects should offer low cost, reliable, and impartial data collection.

The likelihood of successful implementation of Automatic OEE Monitoring across an organisation can be significantly increased through the effective use of low-cost pilot implementations.

 

If you have any queries or would like further information, please do not hesitate to contact [email protected]