The @MATEC Archives

Volume 1, Number 4 Overall Equipment Effectiveness
By:  John Fowler, Ph.D.
Arizona State University
@matec.gif (37291 bytes)

According to the Semiconductor Industry Association's 1997 National Technology Roadmap for Semiconductors (Semiconductor Industry Association, 1997), the semiconductor industry has historically maintained 25-30% annual productivity growth despite factory costs that have increased at about 20% per year (see Figure 1).  This productivity improvement has come primarily from four sources (see Table 1).  While the historic contribution of feature size reductions is likely to continue, the historic contributions of wafer size increases and yield improvements are likely to decrease in the future.  Wafer size increases may still be necessary in the future, but the increased costs of raw wafers will reduce the historic derived benefit.  Probe yield increases become more and more difficult because yields are already quite high.  Therefore, in order to stay on the productivity growth curve, it will be necessary to significantly improve the productivity of semiconductor manufacturing operations.  Since the cost of semiconductor manufacturing equipment will reach about 80% of total factory costs by 2000, most of the improvement will need to come from increased equipment productivity. 


Overall Equipment Effectiveness (OEE) is a comprehensive metric that indicates the relative productivity of a piece of equipment compared to its theoretical performance.  While this metric can be determined for all equipment in the factory, it is particularly important for bottleneck equipment as it identifies areas critical for the improvement of equipment productivity.  OEE takes is formally defined as:

OEE = Availability   x Performance Efficiency x Rate of Quality


The overall OEE metric can also take on values between zero and one.  The first component is the availability of the equipment.  It measures the percent of time that the tool can be used to process wafers.  The availability is the equipment uptime divided by total time as defined in Figure 2.

Performance efficiency is the percentage of available time that the equipment is producing product at its theoretical speed for individual products.  It measures speed losses (e.g., inefficient batching, machines jams) and losses associated with idling equipment (e.g. wait for WIP, wait for operator) or minor undocumented stoppages or assists.

The final component of OEE is the Rate of Quality.  It incorporates losses due to misprocessing and breakage.  It is calculated by determining the percent of the total output in wafers (i.e. all wafers including production, engineering, rework and scrap wafers) that is good.

Calculating OEE

Consider data for the following simple situation.

Total Time  = 168 hr Nonscheduled Time = 0 hr
Unscheduled Downtime = 8 hr Scheduled Downtime = 28 hr
Engineering State =  25 hr Standby State = 7 hr
Productive State = 100 hours
Theoretical Processing Time  =  2 min/wafer* Wafers Processed = 2400 wafers
Rework Wafers = 50 wafers Scrap Wafers = 25 wafers
* 30 wafers/hr

The three components of OEE are determined and then multiplied together to determine the overall OEE which in this case = 46.1%

SEMATECH (1995), which can be accessed at /public/docubase/summary/2745agen.htm, provides a more complicated example with two processes.


TOTAL Productive Maintenance

OEE is the key metric of a Total Productive Maintenance (TPM) program. According to Nakajima (1989) "TPM is a plant improvement methodology, which enables continuous and rapid improvement of the manufacturing process through the use of employee involvement, employee empowerment and closed-loop measurement of results.  In a TPM program, teams comprised of operators, maintenance technicians, engineers and equipment suppliers are formed to improve the productivity of a key piece of equipment in the factory.  The goals of a TPM program include: 1) Reduce manufacturing costs, 2) Maximize the effective use of plant equipment (increase OEE), 3) Increase the skills of the operations and maintenance personnel and 4) Improve employee morale.


(TPM) was first used by Nippondenso (Automotive Component Supplier) in late 60s and Seiichi Nakajima of the Japan Institute of Plant Maintenance (JIPM) began spreading TPM in Japan in the 70s. In the late 80s, TPM began to spread outside of Japan and became known in semiconductor industry in 90s. SEMATECH has become actively involved in spreading the TPM philosophy and has actually changed the scope of TPM to be Total Productive Manufacturing. For more information on TPM see Shirose (1992) or



The primary benefit of the OEE metric is to identify areas for equipment productivity improvement, but it can also be used for benchmarking. It is ideally suited to compare identical tools within a given factory, but can also be used (cautiously) to compare the performance of similar tools across factories. In the latter case, one must be careful to consider the operating conditions (e.g. machine loading, batching policies, etc.) and the value used for Theoretical Processing Time. 

The 1997 Technology Roadmap indicates that OEE for bottleneck tools in today's factories is 65% and the average tool OEE is 45%.  They project that these numbers need to rise to 92% for bottleneck tools and an average of 80% for all tools by the year 2012.  These needs can only be met by operators, maintenance technicians, engineers, and equipment suppliers all working together to improve the equipment productivity.


Nakajima, S., Introduction to TPM, Productivity Press, Portland, OR, 1988

SEMATECH, Overall Equipment Effectiveness (OEE) Guidebook Revision 1.0, Report No. 95032745A-GEN, Austin, TX, 1995.

SEMI, E10-96 Standard for Definition and Measurement of Equipment Reliability, Availability, and Maintainability (RAM), 1996.

Semiconductor Industry Association, The National Technology Roadmap for Semiconductors, 1997.

Shirose, K., TPM for Workshop Leaders, Productivity Press, Portland, OR, 1992.

For more information see:

John W. Fowler is an Assistant Professor in the Industrial and Management Systems Engineering Department at Arizona State University.  Prior to his current position, he was a Senior Member of Technical Staff in the Modeling, CAD and Statistical Methods Division of SEMATECH.  He received his Ph.D. in Industrial Engineering from Texas A&M University and spent the last 1.5 years of his doctoral studies as an intern at Advanced Micro Devices.  His research interests include modeling, analysis and control of manufacturing (especially semiconductor) systems.  John is an avid jogger, a soccer enthusiast, and likes a wide variety of music types from classical to folk.