What is OEE and how can it be applied in extrusion?

Productivity in production is an important factor for the profitability of extrusion companies at all times. Whether it is a matter of achieving the highest possible output quantities in times of high economic activity or of reducing costs in times of difficult economic conditions.

In recent years, the OEE (Overall Equipment Effectiveness) has become established as a key figure for measuring productivity in manufacturing. OEE is composed of three factors: availability, performance and quality.

The losses that occur in the individual components of the OEE are also referred to as availability losses, performance losses and quality losses.

The OEE ratio puts pure productive time in relation to total time:

💡 OEE von 100%

Es wird die gesamte Zeit mit maximaler Geschwindigkeit und ohne Ausschuss gefertigt.

Originally, OEE originated in discontinuous manufacturing, but over time it has been adopted for other manufacturing processes. The success of OEE can be attributed to its ease of interpretation and all-encompassing evaluation of production in a single metric.

In order to uncover concrete improvements, availability, performance and quality losses are broken down even more finely into the “6 Big Losses of Lean Philosophy”.

In the extrusion industry however, we calculate only “5 Big Losses”, since short stops (of less than one minute) cannot occur in continuous production by design. Reasons must also be documented for each lot/loss, and this is ideally done by the workers. Today, feedback from MES/PDA systems is used for this purpose.

Identify the biggest losses with the "Biggest Loss" analysis

Reasons for unplanned downtime can be, for example, demolition, lack of personnel, lack of material, etc. With the help of the “Biggest Loss” analysis, the causes of losses that lead to the largest losses in total are identified. This identifies the biggest levers for improvement.

With the help of the OEE, the progress of the improvement can then be tracked continuously. This is where the advantage of the OEE comes into play.

💡 Egal, welche Verlustkategorie verbessert wird, der Effekt zeigt sich im OEE. Das Management hat damit ein gutes Werkzeug, um Ziele zu setzen und den Fortschritt zu messen.

A correct data basis is the prerequisite for the implementation of the OEE

The resilience of the OEE metric is very important because it is a key control element for improving production processes. Therefore, the OEE should reflect the actual manufactured reality. A correct and solid data basis is a basic prerequisite for this.

Against this background, it is all the more surprising that most OEE implementations in the extrusion industry are based on manual data collection and calculation.

🚨 Im Folgenden stellen wir eine automatisierte OEE-Berechnung, zugeschnitten auf die Extrusion und basierend auf kontinuierlich erfassten Maschinendaten, vor.
Erklärtes Ziel ist es einen OEE zu erhalten, welcher die gefertigte Realität für die Extrusion automatisiert erfasst.

The 3 biggest sources of error in OEE calculation

The 3 main problems in OEE recording for extrusion are:

  1. Machine-based performance consideration
  2. Manual recording of downtime
  3. Manual material recording for calculation of scrap

In the following we will show how these problems can be solved.

1. Machine-based calculation of the performance component

A reference value – the maximum throughput – is needed to determine the performance component of the OEE.

“Maximum machine output” as a source of error

In extrusion, the maximum machine output (machine-based calculation) is typically used for this purpose. However, this approach neglects the fact that different products can achieve different throughputs.

→ The solution: Use the production history to calculate the maximum throughput on a product-specific basis.

Therefore, instead of the maximum machine output we recommend using the maximum throughput based on the manufacturing history as the reference value (product-based calculation), which has been achieved stably for the respective product. This throughput value is then stored as MDS (Maximum Demonstrated Speed) for the respective product.

💡 Die Ermittlung des MDS wird bei ENLYZE vollautomatisch im Hintergrund durchgeführt und für die Berechnung der Leistungskomponente verwendet.

Applied to discontinuous production, the MDS corresponds to the ideal cycle time, which is used as a reference value for discontinuous processes. In our opinion, only this product-related reference makes it possible to make meaningful statements regarding the performance component and thus the OEE.

The approach of choosing heuristically defined target values as reference values helps to a certain extent, but here, too, the constant comparison with reality is missing, since the reference values change continuously.

Clear differences in precision: machine-related vs. product-related reference value

Figure 3 clearly shows the differences between a machine-based (conventional consideration, left) and a product-based (ENLYZE consideration, right) calculation.

In the example shown above, product A can be produced at a maximum of 340 kg/h and product B only at 275kg/h. The max. machine capacity is 350 kg/h. In the conventional approach, the max. machine capacity of 350 kg/h is used as a reference for all products. ENLYZE uses the product-specific reference values (Product A 340 kg/h; Product B 275 kg/h).

Machine-based calculation (conventional):

In a machine-based consideration, order FA1 with product A performs significantly better with a performance factor of 88.6% than FA2 with product B with a performance factor of 77.1%.

Product-based calculation (ENLYZE):

However, in a product-based consideration, FA2 with product B performs significantly better with 98.2% than FA1 with product A with 91.1% . This shows how large a bias in the performance factor can be with a machine-based compared to a product-based approach.

Inaccuracies of up to 20% with incorrect OEE calculation:

The difference is clearly evident: the differences in the performance component show a difference of over 20 percentage points for FA2. It can be seen that a product-based calculation is worthwhile in order to compare the different products fairly.

2. Manual recording of downtimes is a major source of error.

Another major source of error in the calculation of OEE is the inaccurate recording of downtimes and their duration, which leads to an incorrect availability factor.

Typically, downtimes are recorded today via manual bookings by the operator in the MES or PDA system. However, the downtimes and in particular the duration of the downtime are subject to inaccuracies due to the manual bookings.

The reason for the inaccuracies is that in the event of a shutdown, the operator first wants to eliminate the problem and get the machine running again. The booking of the standstill and the duration of the standstill usually takes place only afterwards and is only roughly estimated in terms of time. In some cases, paper-based shift logs are still used today, which are only filled in at the end of the shift. Here, entire downtimes are often forgotten.

For the calculation of the availability factor of the total loss, the sum of the downtimes is set in relation to the total time. Inaccuracies in the downtimes thus add up and can lead to considerable inaccuracies in the availability factor.

Record downtimes automatically

We therefore believe that the recording of downtimes should be automated, without manual bookings, digitally and without pen and paper.

💡 Idealerweise sollten Stillstände digital und automatisiert erfasst werden und auf Maschinendaten beruhen. Nur dadurch kann eine verlässliche Datengrundlage geschaffen werden.

Example of automated downtime recording:

Shutdowns can be detected, for example, on the basis of the machine throughput. If the machine throughput falls below a certain limit, then the start of a shutdown is detected. As soon as this limit is exceeded again, the end is detected. This ensures, that the reality produced is derived accurately and automatically from the machine data and at the same time relieves the operator as manual bookings are no longer necessary. In addition, the operator can directly correct the problem at the time of the shutdown and can specify the reason for the shutdown afterwards: A win-win situation.

The recorded shutdowns and the reasons for them can then be used in a subsequent “5 Big Losses” analysis to identify the biggest levers for avoiding shutdowns.

3. Inaccurate material recording

After a correct recording of the performance factor via product-based benchmarks as well as an exact and automated time recording of downtimes for the availability factor, the quality factor still remains as the last open component.

To calculate the quality factor of a production order, the scrap quantity must be set in relation to the total plasticized quantity.

The exact determination of the yield often poses no problems in practice. For this purpose, the quantity of material is usually weighed or recorded in linear meters or other quantities and booked into the ERP/MES system. Blocked quantities are subsequently taken into account by QA in these postings. It is more difficult to record the plasticized quantity or the scrap.

The plasticized quantity can usually be determined via the throughput (e.g. from gravimetry). For this purpose, an integral is formed over the throughput, with the limits being the start and the end of the respective booked order. By subtracting the yield from the plasticized quantity, the scrap quantity is calculated. The quality factor can then be calculated.


The objective was to present an OEE determination method that reflects the manufactured reality in extrusion in order to make data-based, targeted decisions to increase productivity.

Much of the data collection can be automated. However, the start and end of a job must still be done manually. However, errors caused by manual postings are drastically reduced.

Accurate OEE calculation is ensured by:

  1. Product-based reference values for the performance calculation.
  2. Accurate and automated recording of downtimes based on machine data
  3. The evaluation of the quality factor, based on the automatically recorded plasticized quantity

In this way the OEE represents the actual manufactured reality and can be used meaningfully to control production.

In addition, there are further advantages to be gained from automated recording of the OEE:

  • Continuous availability of key figures without manual input need
  • Building trust in the database
  • Significant reduction of the effort required to maintain the system

ℹ️ ENLYZE hat diese Art der OEE-Ermittlung mit fünf Extrusionsunternehmen über das letzte Jahr hinweg in der betrieblichen Praxis erprobt und konstant weiterentwickelt.

Inzwischen nutzen knapp 15 Kunden erfolgreich die OEE-Tools von ENLYZE. Im Schnitt konnten die Unternehmen mit diesem datengetriebenen Ansatz 3,6 % Produktivität in den ersten 3 Monaten heben und langfristig Ihre Produktivität um 5-20 % steigern.

Neben der Berechnungen des OEEs liefert ENLYZE auch die passenden Analyse Tools, um bei Produktivitätsverlusten Ursachen-Analysen durchzuführen. Die ersten Produktfeatures des ENLYZE Shop Floor BI richten sich dabei an die Produktions-, Werks- und Betriebsleitung sowie an Lean- und Operational Excellence Manager.
Scroll to Top