Mining the Hidden Factory

Like Lean Manufacturing and Six Sigma, Overall Equipment Effectiveness (OEE) is a key tool for achieving operational excellence. Although the pharmaceutical industry is five to 12 years behind discrete manufacturing in applying this concept, more pharmaceutical companies, including AstraZeneca and Merck, are including OEE as part of their operational excellence toolkits and discovering how it can help quantify efficiency and improve overall operations.

This article will present a brief overview of OEE and show how the concepts behind it can help drive manufacturing excellence at your operations. It will also highlight the critical link between OEE data and business metrics, notably Income from Operations (IFO). This connection is essential in communicating to management, and in helping you develop the most effective strategy for achieving operational excellence at your facility.

OEE’s roots can be found in Total Productive Maintenance, a term coined by Seiichi Nakajima [1] that quantifies all losses or deviations from perfect manufacturing. TPM is completely in line with Lean Manufacturing and the Toyota production system.

Simply defined, OEE is: The amount of ideal production time relative to the actual scheduled time used, or the amount of good product produced relative to the amount of total product that could have been made with 100% first-pass yield at the ideal rate for the actual scheduled time used.

Also important to OEE is Total Effective Equipment Performance (TEEP), or the amount of ideal production time relative to the actual calendar time used, or the amount of good product produced relative to the amount of total product that could have been made with 100% first-pass yield at the ideal rate for the actual calendar period used. This metric is used to determine overall asset usage before investing capital for capacity expansion. A simple “back of the envelope” calculation can be done, and usually the TEEP number is significantly lower than the current OEE value, particularly for pharmaceutical companies that work a relatively low number of shifts and weekends when compared with other industries.

Figure 1 may be helpful in allowing you to visualize OEE and TEEP. The gold portion of the chart represents the amount of “perfect production” time, or “Theoretical Factory” time, arrived at by dividing the total amount of good product produced by the ideal speed rate of units per time increment.

Losses generally take one of the following forms:

• Planned —
Unlike TEEP, OEE excludes all planned downtime. Planned downtime would include major shutdowns, periods during which there are no orders, planned holidays, weekends, shifts not worked, planned lunches and breaks and meetings, experiment time, and downtime for completing orders early.


• Operational —
These losses occur when the equipment is not running, such as during changeovers, pit stop maintenance, process stops for cleaning and unplanned equipment problems, quality problems, lack of manpower, lack of information or lack of supplies. Operational losses can be planned or unplanned, and data collection is critical in eliminating unplanned losses and mitigating planned downtime. Automated data collection can be particularly useful, especially data organized based on product run.


• Speed —
These types of losses occur when operations do not run at the ideal speed, or the highest accredited speed or equipment design speed limits. (This factor must be 1.0 or less. If it is greater than 1.0, it means that a new ideal speed has been established.) Minor equipment stops and jams accumulate as speed losses. So do instances where runtime speed is reduced.

If the specific equipment system being reviewed is not the bottleneck of the product flow, then the ideal speed rate could be defined as the desired rate to feed the bottleneck. OEE is then measured against desired speed with the understanding that the maximum speed factor is 1.0. All events that require higher than desired speed to meet order rate should be recorded for analysis and improvement opportunity.


• Quality —
This type of loss results when any produced units are not transferred on the first pass. Rework has zero OEE value until it is corrected and transferred.

OEE is important for operations, such as pharmaceutical manufacturing and packaging, that run fewer than 24 hours per day, or are not sold out. TEEP is the important metric for 24/7 or “sold out” operations.

Recently, demonstrated world-class OEE levels were determined for the pharmaceutical industry (Pharmaceutical Manufacturing, January 2006, p. 11). Levels were determined for different business conditions, depending on the number of products produced by the equipment system. The following was information combined from Granulation, Compression, Film Coating, C/E, and hybrid operations:

Changeover reduction

Since the drug industry is highly regulated and extremely safety-conscious, SKU changeovers are critical events. For even the most efficient pharmaceutical companies, the best demonstrated changeover time is still close to 2 hours. Compare that to the times achieved by other industries, some of which have reduced die changes from over 12 hours to less than 10 minutes. Changeover time reduction is likely to remain a top goal for pharmaceutical companies.

Pharmaceutical operations are unique, in that both discrete and batch processes are used to manufacture products in their final packaged form. Discrete manufacturing is usually automated, and the ideal speed is the design speed of the equipment (i.e., a discrete process such as a bottle filling operation is designed to fill a specified number of units per minute and the process usually occurs as the bottle is moving through a defined distance along a curved track).

Drug companies embarking on an OEE program typically start with their discrete processes, probably because they are better defined and fairly straightforward for measuring and analyzing event times and speeds. However, OEE can also be applied to batch operations, where it can allow significant improvement to be made.

With batch processes, it is critical to define the portions of the transformation process that will add the most value. Standard operating procedures (SOPs) must be developed for every portion of the cycle.

It is also essential to quantify the ideal rate of operation for each portion of the process, which is usually expressed as the best time that could be achieved by operating equipment that functioned “like new.” The figure that results is used as a baseline against which to measure losses due to fouled heat exchangers, pump degradation or clogged pipes. Specific “best practice” heat ramps and process curves should be defined, as well as proper mixing of ingredients, dwell times and transfer processes.

Once these steps have been taken, the first, and usually the biggest, improvement that results is that cycles will be executed exactly the same way every time, on every shift, using established best practices and quality ingredients.

The next step is to reduce the time required for non-value-adding operations such as cleaning, decontamination and start preparations including inspections and information downloads, to drive them to “pit stop” conditions. Changeover practices should be sped up as much as possible, and all actions examined to see if they can be done outside of the “critical path.” Examples might include using alternating parts and cleaning off-line. Even for value-adding portions of the critical path, parallel processing can be done — for example, preheating or cooling equipment or ingredients while a production line is running.

Using statistical process control (SPC) instead of testing or implementing process analytical technologies (PAT) would be ways to eliminate losses in time and material due to testing. These practices could also minimize materials left over after transfer and mitigate losses from pumping, tanks or production lines. A good calibration system will be needed, in order to confirm that all measurements are accurate and that they are providing good information.

Defining the quality factor

Some organizations suffer from variable batch process yields and argue that the “quality factor,” or good quantity versus total quantity produced, would be impossible to define.

In such cases, carefully designed experiments will make the process more scientific, and demonstrate the sensitivity of all parts of the process and materials. At the very least, careful records should be used to define the “best ever” yield, to set the standard for demonstrated capability of current operations.

Another way to visualize the OEE chart (Figure 1, below) is to look at the chart vertically, with the gold portion at the bottom. When viewed this way, the height of the chart is the “capacity ceiling,” and the gold portion represents current capability.

Next, draw a line for world-class TEEP, which would be approximately two percent lower than the world-class level for OEE, or at least 81%, or about 295 days, for the pharmaceutical industry.

The difference between that number and the number of Theoretical Factory days that your organization would need in order to produce at an ideal rate represents the effectiveness loss, or the profits that your organization loses every year.