Overall Weighting Equipment Effectiveness R. Wudhikarn College of Arts, Media and Technology, Chiang Mai University, Thailand ([email protected])

Abstract – This research is designed to improve the original Overall Equipment Effectiveness (OEE). The OEE is the process, in which acquired to specify an equivalent weight setting of every single element, even though; each concerning losses are totally different. Hence, the study proposes a simpler weight setting method, so called the Rank-Order Centroid (ROC), to identify dissimilarity in weighting each OEE element. The ROC methodology is easier to determine the weight than the existing weighted OEE method, which is based on an analytical hierarchy process. This newly calculating methodology is the Overall Weighting Equipment Effectiveness (OWEE). It is presented and also implemented in a fiber cement roof manufacturer. The result granted from OWEE, however, is different from those of the original OEE and of the existing weighted OEE.

an indicator of the process improvement activities within a manufacturing environment [2]. Therefore, several studies [3 - 5] have implemented the OEE and, accordingly, major improving outcomes can be determined. However, the OEE has some flaws; especially, the weighting of each element is equivalent, whilst, their losses are totally different. For example, a quality rate is composed of qualitative losses but availability rate associates with the time collapsed. Therefore, many studies, including this study, have attempted to improve this weakness. This study proposes to improve weaknesses of the OEE by developing an existing calculating methodology. The theoretical concept of the OEE measurement is, also, taking into consideration, thence the problems and the development of the OEE are examined and discussed with the regarding literatures. Later, the proposed methodology is presented, a case study application is, then, analyzed and, finally, the potential of the new calculating method is explored. This methodology has aimed to rectify the weight setting of the OEE’s elements.

Keywords – Overall equipment effectiveness, Total preventive maintenance, Rank order centroid, Normalization, Analytical hierarchy process

I. INTRODUCTION Several business companies are now facing an intense competitive condition. It has pushed these companies to enhance their production availability, performance efficiency and also product quality, in order to take advantage over their contenders. The high demand, also, plays role in stimulating firms to increase their production capacity. In order to do so, most of labors are being replaced by automatic machines. They are believed to have higher reliability, faster production rate, more precision and contain lower operating cost than human, in general. This condition would be possible, if and only if; these equipments are able to perform with a higher effectiveness and efficiency. These excellent performances, however, also required an appropriate management system for monitoring its function. Total Preventive Maintenance (TPM) is one of those broadly applied management systems, as it is employed to strengthen the manufacturing performance and to achieve the world-class performance [1]. Still, these management systems require an appropriate information system to evaluate their operating performances. The information gathered from these operating machines is mandatory for sustaining the organizations. One of the TPM’s achievements is to enhance the equipment efficiency. The calculation is made by using Overall Equipment Effectiveness. The OEE method is based on three main elements of performance consisting availability, performance efficiency and qualitative aspect. The OEE is not only defined as an operational measure, but also employed as

978-1-4244-8503-1/10/$26.00 ©2010 IEEE

II. METHODOLOGY A. Overall Equipment Effectiveness In 1988, Nakajima [6] proposed the OEE as a tool for assessing the success to his proposed TPM philosophy. According to Nakajima, the OEE is based on three main aspects; each element concerns with different losses (see Table I). TABLE I PERFORMANCE ASPECTS OF OEE AND RELATING LOSSES Performance aspects 1. Availability Rate

Relating losses - Equipment failure/breakdown losses - Set-up and adjustment losses

2. Performance Efficiency

- Idling and minor stoppage losses

3. Quality Rate

- Defect and rework losses

- Reduced speed losses

- Start-up losses

From the Table I, these losses are defined as the “six big losses” and their definitions are described as follows: 1) Equipment failure/breakdown losses: These losses are categorized due to time losses (reduced productivity), and quantity losses (occurrence of defective products), which are related to sporadic/chronic failures. The sporadic failures happen when the changes occur in

23

Proceedings of the 2010 IEEE IEEM

some conditions (jigs/tools, work method, and equipment’s state). These require a mean of measurement to revert them back to their original conditions. The chronic failures occur when there are some hidden defects in the operating equipments. They are uncorrectable, even if various countermeasures are exploited. 2) Set-up and adjustment losses: These represent the time losses in between the endings of the production and the time losses due to adjusting of equipments to meet the requirement of a new item production. A setting up means a series of operations from the removal of jigs and fixtures to the end of production, clearing up and cleaning, until the preparation of jigs2tools and the necessary metal fixtures for the next product, plus their attachment, adjustment, trial processing, readjustment, measurement, production, and ,finally, the ability to produce excellent products. These first two losses are considered as the time losses, which are significant for calculating the availability rate of the equipment. 3) Idling and minor stoppage losses: These losses take place when the production is interrupted by a temporarily malfunction or when a machine is idling. For instance, the idling and minor stoppages caused by the malfunctioning of conveyors and blockages of material in a pipe. 4) Reduced speed losses: These losses are according to the delay in equipment speed. To specify; (i) the actual speed is slower than the design speed; (ii) the design speed is slower than the present technological standards or the desirable condition. These third and fourth losses are the speed losses, which determine the performance efficiency of the equipment. 5) Defect and rework losses: These include the volume and time losses due to defect and rework (disposal defects), financial losses due to a product downgrading, and repairing time losses for defective products. 6) Start-up losses: Start-up losses are time and volume losses. For instance, (i) Start-up after periodic repair; (ii) Start-up after suspension (long-time stoppage); (iii) Start-up after holidays and (iv) Start-up after breaks These last two losses are regarded as the quality losses. They affect the quality rate of the equipment directly. Normally, the OEE is calculated using these six big losses. These losses are the functions of Availability Rate (A), Performance Efficiency (P) and Quality Rate (Q) which can be determined as in the following equation:

a ratio between the actual operating time and the loading time. (2) The net operating time is the time durable, in which the equipments are processing at a standard production rate. To calculate the net operating time, the performance time losses are subtracted from the operating time. The performance time losses consist of normal production losses (production rate reduction due to start-up, shutdown, and changeover) and abnormal production losses (production rate reductions due to abnormalities). On the other hand, the net operating time is the processed amount multiplied with the actual cycle time. Where the processed amount refers to the number of items processed per time period (day, month or etc.), and the operating. (3) The defect amount represents the number of items rejected due to quality defects, and requires a rework or, otherwise, would become scrapped. Combining (1) to (3), the OEE for the given equipment operation is computed using: OEE = A x P x Q

(4)

As presented, the OEE depends on the availability rate, the performance efficiency and the quality rate. Therefore, the OEE is in direct proportion to all these three elements. Rising in the availability rate reduces buffer inventories needed to protect a downstream production from breaking down and, at the same time, increases its effective capacity. The reduced buffer inventories lead to a decreasing in lead times, since the jobs are not waiting as long as in the queues. This capability of shortening the lead-times improves the firm’s competitive position; in terms of delivery and flexibility, since it is easier to deliver multiple products or variations of products in a shorter lead time. The reduction needed for the buffer inventory minimizes the inventory costs directly and, later, results in increasing of the effective capacity. This allows more throughputs and lowers the cost per unit. An escalating in the performance efficiency, also, lowers the needs for the buffer inventories, together with enhances the effective capacity. This reinforces the benefits gained from an increasing in equipment availability. An upgrading in the rate of quality products can be implied that there is less scrap and rework, which is not only reduces the costs, but also yields a higher rate of quality [7, 8]. Although the OEE seems to be a complete performance measurement indicator, still, it requires a proper modification. For instance, an existing research had showed the difference in element weights of the OEE [9]. The conclusion of this study was; the traditional means of evaluating the maintenance management

(1) , where the loading time is the planned time available per time period (day, month, etc.) for production operations, and the operating time is calculated from the loading time subtracted by the time of equipment failures, set-up and adjustment requirements, exchange of dies and other fixtures and etc. The availability, in this case, is defined as

24

Proceedings of the 2010 IEEE IEEM

systems could not yield higher capital productivity. Factors relating to the OEE are not equally important in these aspects; however, the different in weights should be taken into account. Furthermore, there is another research [10] engaged with the calculating method, which proposed; how to calculate an increasing in profits or a decreasing in costs from an increasing percentage of the OEE. Hence, all the outcomes are presented in term of the monetary unit allowing easier ranking of the problem priorities. These studies have put efforts on improving the OEE’s weaknesses to prevent a wrong decision making. They are, however, inappropriately developed, thus, a new indicator has been proposed [11]. This modified method is called an “overall equipment cost loss indicator”. These last two mentioned methods, nevertheless, require the operating information accompanied with the financial information, which put these indicators into a complicated shop floor. As a conclusion, the OEE or the weighted OEE is more suitable for the operational process.

reasonable than the original OEE, as its loss in each element is different. Though, the weight setting should not be equivalent. Applying of the analytical hierarchy process with the simple weight setting is considered to be more complicated for the decision makers; hence, this research proposes another alternation of the simplifying appropriate weighting method. C. Rank-order centroid (ROC) The Rank-order centroid (ROC) is proposed by Edwards and Barron [12]. This technique transforms the swing ranks into the swing weights. Moreover, the ROC weights define the best alternative 75 up to 90% of the time based upon a set of the true swing weights elicited some other way accurately. The calculating formula of the ROC is defined as in (6) ∑

Where is the rank of the th objective, is the total is the normalized number of the objectives, and approximate ratio scale weight of the ith objective. The ROC method is, also, applied in this research and the weighting factors for each element are determined by interviewing a top manager who has a total authority of the company management.

B. Weighting of the OEE The traditional OEE specifies the weighting of each element equivalently. This designation, still, is inappropriate. These elementary losses are totally different, as the availability rate is regarded to the time losses, the performance efficiency concerns with the speed losses and the quality rate is defined as the quality losses. So Raouf [9] had proposed a modified OEE. This methodology assigns weights to all the elements using the analytical hierarchy process. Assuming that has a and has a weight of weight of , has a weight of 1 and ∑ 1. In this case, the , where 0 weighted OEE can be calculated as

D. Normalization In general, normalization refers to the comparison of the overall size of category indicators with the reference information [13]. Its main purpose is to resolve the issue of the unit differences. As a result, the normalization aims at obtaining a set of comparable scales, which allow interattribute as well as intra-attribute comparisons [14]. As mentioned, the units and ranges of the time losses, the speed losses and the quality losses are different and, in order to make a comparison, these dissimilarities should be extinguished. The calculating formulas of , and are the ones taking responsible for that, or, to say, to change into a normalized form. These comparative values are used in accompanied with the correlative weight for calculating a summarization of the overall equipment performance.

(5) The numerical example between the original OEE and the weighted OEE are represented in Table II. TABLE II NUMERICAL EXAMPLE OF OEE AND WEIGHTED OEE

M/C

A

P

Q

Original OEE

Weighted OEE

No.1

89.0%

81.4%

98.0%

71.0%

90.9%

No.2

91.2%

79.7%

98.1%

71.3%

90.8%

Weight

0.20

0.30

0.50

-

-

(6)

III. PROPOSED METHODOLOGY As stated prior, this research main goal is to find the way to improve the OEE method. To do so, the newly calculating method combined the traditional OEE approach with the weighted OEE approaches is proposed. First, the OEE elements are still computed as in the original calculation, following (1), (2) and (3) for , and respectively. Then, the weight of each element is specified by using the ROC method as

The Table II depicts a different weight setting for each single element. However, the first problematic machine exhibited a contradiction between the original OEE contrasts and the modified OEE, since, the weighted OEE specified the highest vital at the quality rate element. This method, on the other hand, has proved to be more

25

Proceedings of the 2010 IEEE IEEM

The weighting and the OEE’s element data from the Table III and the Table IV were, then, employed to calculate the traditional OEE and the OWEE. Moreover, Raouf’s OEE was also exploited as a comparison mean for both methods. The outcomes are exhibited in Table V.

illustrated in (6). Afterward, the overall weighting equipment effectiveness (OWEE) is calculated utilized the following formula. (7)

OWEE

TABLE V NUMERICAL RESULTS FROM THREE METHODS

, where; is the weight of the availability rate element, is the weight of the performance efficiency element is the weight of the quality rate element. and The OWEE can be simply applied to any company already implementing the OEE. Anyway, firstly, the weight must be identified by an authorized person.

IV. OWEE MODEL APPLICATION In this section, the application of the OWEE model previously discussed is exemplified through a case study. In this newly improved calculation method, the production data such as the processed products, the good products, the loading time, the operating time, and etc. are collected, as the weight setting is mandatory. In this case study, a managing director was interviewed for prioritizing the availability rate, the performance efficiency and the quality rate element. The outcome of the order of magnitude is demonstrated in the quality rate, the performance efficiency and the availability rate respectively. Then, the prioritized orders were applied for calculating weight by using the ROC method resulting in Table III.

Ranking

A

3

P

2

Q

1

Numerical calculation 1⁄3 ⁄3 1⁄2 1

1⁄3 ⁄3

1⁄2

1⁄3 ⁄3

Original OEE

Raouf’s OEE

OWEE

ST1

74.29%

95.14%

93.85%

ST4

73.80%

94.57%

93.32%

ST5

73.82%

94.02%

92.89%

CC3

75.75%

94.97%

93.81%

CC4

76.91%

94.62%

93.65%

CC5

77.17%

95.08%

94.01%

V. DISCUSSION From the Table V, the results from those three different methods have been calculated and the outcomes are presented in percentage. The machine ST4 shows the lowest traditional OEE, even though; for the other remaining methods, the machine ST5 comprised a lower score. This irrelevant outcome, however, is expected, as the OEE normally identifies an equivalent weight in each single element, on the other hand, both of the modified methods do not. Therefore, the original OEE possibly gives a different result from other weighting methods. Later, the problematic machine can be ranked by alternatively using the original OEE, the Raouf’s OEE and the OWEE. The results are displayed in Table VI.

TABLE III WEIGHT SETTING BY ROC METHOD Element

Machine

Weight

TABLE VI MACHINE PRIORITY BY ALTERNATIVE METHODS

0.11 0.28 0.61

Machine

*Ranking by Original OEE

Raouf’s OEE

OWEE

The OEE element data of the selected company, a fiber cement roof manufacturer, between June and December 2009 are depicted as illustrated in Table IV.

ST1

3

6

5

ST4

1

2

2

ST5

2

1

1

TABLE IV

CC3

4

4

4

OVERALL EQUIPMENT EFFECTIVENESS OF MACHINES

CC4

6

3

3

CC5

5

5

6

Machine

A

P

Q

ST1

88.98%

84.21%

99.15%

ST4

91.58%

81.27%

99.16%

ST5

92.41%

81.26%

98.31%

CC3

92.03%

83.12%

99.03%

CC4

93.19%

84.14%

98.09%

CC5

93.52%

83.36%

98.99%

*

Less number means high priority to improve

From the Table VI, the problematic machines ranked by the OWEE method are almost similar to the Raouf’s OEE. The OWEE’s outcome is mostly dissimilar from the OEE’s, yet, only in the fourth order. Anyway, it shows slightly differences from those of Raouf’s OEE in the fifth and the sixth order. From the newly proposed method, the machine ST5 is the first improved machine relating to the Raouf’s OEE. This similarity is the result of the and the

26

Proceedings of the 2010 IEEE IEEM

[4] M. Lesshammar, “Evaluation and improvement of manufacturing performance measurement systems – the role of OEE”, International Journal of Operations and Production Management, vol. 19, no. 1, pp. 55-78, (1999). [5] R. Wudhikarn and W. Manopiniwes, “Autonomous maintenance using total productive maintenance approach: A case study of synthetic wood plank factory”, Technology Innovation & Industrial Management Conference, TIIM2010, Pattaya, Thailand, in press. [6] S. Nakajima, “Introduction to TPM”, Productivity Press, Cambridge, MA., 1988. [7] M. Lesshammar, “Evaluation and improvement of manufacturing performance measurement systems – the role of OEE”, International Journal of Operations and Production Management, vol. 19, no. 1, pp. 55-78, (1999). [8] L.D. Frendall, J.W. Patterson and W.J. Kneedy, “Maintenance modeling its strategic impact”, Journal of Managerial Issues, vol. 9, no. 4, pp. 440-448, 1997. [9] A. Raouf, “Improving capital productivity through maintenance”, International Journal of Operations and Production Management, vol. 14, no. 7, pp. 44–52, 1994. [10] O. Kwon and H. Lee, “Calculation methodology for contributive managerial effect by OEE as a result of TPM activities”, Journal of Quality in Maintenance Engineering, vol. 10, no. 4, pp.263-272, 2004. [11] R. Wudhikarn, C. Smithikul and W. Manopiniwes, “Developing overall equipment cost loss indicator”, in Proc. 6th Conf. of Digital Enterprise Technology, DET2009, Hong Kong, Hong Kong, pp. 557-567. [12] F.H. Barron and B.E. Barrett, “Decision quality using ranked attribute weights”, Management Science, vol. 42, no. 11, pp. 1515–1523, 1996. [13] J. Guinée, “Handbook in Life Cycle Assessment”, Operational Guide to the ISO Standars, Kluwer Academic, Dordrecht, 2002. [14] C.L. Hwang and K. Yoon, “Multiple attribute decision making: Methods and applications: a state-of-the-art survey”, Springer-Verlag, Berlin and New York, 1981.

elements, which contain the lowest performance. The , however, is the highest weight element, due to its lowest score. For the differences between the OWEE and the original OEE, the ST4 is identified as the first problematic machine for the OEE traditional method. As the OWEE identified the highest weight setting on the , then the overall result mostly depends on this element. As for the flaws of this study, this numerical case study contains only a few sets of machines and, plus, the performance results are also nearly similar. For a better alternative of the further developed study; it is suggested to have more machines and also a vast variation in capability. Still, at this moment, there is no completely optimal method for every circumstance. The methodology is highly depending on, which is the most appropriated mean to specify a problem. Finally and importantly, everything is depending on the decision maker, solely. VI. CONCLUSION The performance measurement for operating a machine, together with an appropriate decision from an authorized manager, is crucial for sustaining a business organization. Therefore, it is mandatory to establish a proper mean of measurement. In addition, accuracy in the performance measurement is essential to improve and is the key to success in a business goal. One of the most important and widely used metrics of performance in the manufacturing is the OEE, especially, for firms already applying the TPM. The original OEE method does not appropriately prioritize the problematic equipments. It specifies an equivalent weight in each element, then, the weighted OEE method is presented and the analytical hierarchy process has been applied for setting the weight. This methodology is, however, complicated for a decision maker, especially, those who have not priory became accustomed to this method. To avoid these conflicts, this research has proposed a simpler mean of weighting method and, at the same time, provides a summarized calculation, known as the OWEE. Finally, the implemented case study is also discussed. It is proved that the prioritized problematic machine demonstrates differences from the original OEE and the Raouf’s OEE. REFERENCES [1] K.E. McKone, R.G. Schroeder and K.O. Cua, “The impact of total productive maintenance practices on manufacturing performance”, Journal of Operations Management, vol. 19, no. 1, pp. 39–58, 2001. [2] B. Dal, P. Tugwell and R. Greatbanks, “Overall equipment effectiveness as a measure of operational improvement”, International Journal of Operations & Production Management, vol. 20, no. 12, pp. 1488–1520, 2000. [3] D. Kotze, “Consistency, accuracy lead to maximum OEE benefits”, TPM Newsletter, vol. 4, no. 2, 1993.

27