Art

Constructing the OGE for promoting tool group productivity in semiconductor manufacturing

Description
Constructing the OGE for promoting tool group productivity in semiconductor manufacturing
Categories
Published
of 26
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  This paper has been published in  International Journal of Production Research . Reference to this paper should be made as follows: Chien, C.-F., Chen, H.-K., W, J.-Z., Hu, C.-H. (2007), “Constructing the OGE for promoting tool group productivity in semiconductor manufacturing,”  International Journal of Production Research , 45(3):509-524. (http://dx.doi.org/10.1080/00207540600792515) 1 Construct a Group OEE for Promoting Tool Group Productivity in Semiconductor Manufacturing Chen-Fu Chien * , Hsin-Kai Chen, Jei-Zheng Wu, and Chih-Han Hu Department of Industrial Engineering and Engineering Management, National Tsing Hua University, 101 Section 2 Kuang Fu Road, Hsinchu 30013, TAIWAN *email: cfchiena@tsmc.com Abstract  Overall Equipment Efficiency (OEE) has been proposed as a comprehensive index for individual equipment performance. However, there are limitations for evaluating only OEE of single machine in semiconductor manufacturing. This study aims to propose Overall tool Group Efficiency (OGE) indices to observe the equipment performance at tool group level. In addition, rather than evaluating equipment OEE for a fixed period of observation time, we also proposed a mechanism with statistical efficiency control charts to continuously monitor OGE over time. This study constructed a framework for monitoring equipment performance from critical tool groups to single machines via longitudinal analysis. Then, root-cause analysis of machine statuses is employed to identify possible performance detractors. Following the proposed framework, improvement actions can be thus triggered on those critical machine statuses to promote tool group productivity. We used a case study based on real data from a fab in Taiwan for illustration and concluded this study with discussion of future research directions.  Keywords: Overall equipment efficiency; Tool productivity; Tool group; Semiconductor manufacturing. 1. Introduction Semiconductor fabrication consists of a lengthy and cycling sequence of complex processes by using very expensive equipment. Semiconductor manufacturing is a capital-intensive industry in which reducing cycle time, producing high quality products, on-time delivery, continual reduction of costs and improving efficiency are the most direct and effective ways for maintaining competitive advantages (Peng and Chien 2003). To capture the overall equipment performance for identifying and analyzing hidden performance losses, Nakajima (1988) proposed the overall equipment effectiveness that considers the availability, performance, and  This paper has been published in  International Journal of Production Research . Reference to this paper should be made as follows: Chien, C.-F., Chen, H.-K., W, J.-Z., Hu, C.-H. (2007), “Constructing the OGE for promoting tool group productivity in semiconductor manufacturing,”  International Journal of Production Research , 45(3):509-524. (http://dx.doi.org/10.1080/00207540600792515) 2quality of the equipment. Semiconductor Equipment and Materials International (SEMI) further named the metric as Overall Equipment Efficiency (OEE) (SEMI E079-0200, 2000). As a standard for measurement of equipment productivity, OEE has been widely applied in semiconductor manufacturing. Nevertheless, conflicts and uncertainties often exist among these performance indices such as cycle time, throughput, WIP, utilization and operational efficiency. This study aimed to propose Overall tool Group Efficiency (OGE) indices to observe the equipment performance at tool group level. While most studies focus on the performance analysis for single machine, little research has been done to examine OEE for tool group and evaluate OEE over time. In addition, we also proposed statistical efficiency control charts to continuously monitor OEE over time. Furthermore, this study constructed a framework in which equipment performance can be monitored from critical tool groups to single machines via longitudinal analysis and then root-cause analysis of machine statuses can be employed to identify possible performance detractors. We validated this approach in a fab and the results showed its practical viability. The remainder of this paper is organized as follows. Section 2 reviews the fundamentals of equipment efficiency measurement. Section 3 proposes the indexes for evaluating overall tool group efficiency. Section 4 describes the framework to monitor tool productivity and diagnose loss by OGE scores over time. Section 5 describes a case study in semiconductor manufacturing. Finally, Section 6 concludes with discussion of future research directions. 2. Fundamentals Nakajima (1988) proposed the measure of the overall equipment effectiveness of individual equipment as a metric for evaluating the progress of Total Productive Maintenance (TPM) that integrates the concepts of continuous production improvement and participation of all employees and departments to maximize overall equipment productivity and factory performance. Nakajima (1988) also classified three groups of equipment inefficiencies unfavorable to equipment availability, performance rate, and quality, respectively. In particular, different loss types are identified as follows: (1) equipment failure or breakdown losses and (2) set up or adjustment time losses that will reduce the equipment availability; (3) speed losses and machine idle that will reduce the performance rate; (4) quality loss and yield loss occurred that will reduce the quality rate. Indeed, TPM strives for equipment quality enhancement while Total Quality  This paper has been published in  International Journal of Production Research . Reference to this paper should be made as follows: Chien, C.-F., Chen, H.-K., W, J.-Z., Hu, C.-H. (2007), “Constructing the OGE for promoting tool group productivity in semiconductor manufacturing,”  International Journal of Production Research , 45(3):509-524. (http://dx.doi.org/10.1080/00207540600792515) 3Management (TQM), a predecessor to TPM, is primarily concerned with product quality improvement in terms of qualitative management terms (Leachman 1997). In addition, Statistical Quality Control (SQC) focuses on measuring quality and assuring compliance with specifications and is thus complementary with TQM to ensure product quality that can be measured in various ways of internal quality measures including scrap and rework rates, pressure, temperature and external quality measures referring to customer satisfaction such as return rate and indirect indices derived from sampling, inspection, field service data, customer surveys. Acceptance sampling, process control, and design of experiments are major classes of s aiming at specification conformation. In particular, Statistical Process Control (SPC) is one of SQC tools dealing with measurable quality attributes in manufacturing process and thus can specify process variability sources affecting product quality. Conventional SQC techniques deal with only quality issues via monitoring process data. This approach extended SPC concept to develop control charts for monitoring equipment efficiency metrics to enhance tool productivity. For example, equipment inefficiency might result from speed degrading due to out-of-tune, malfunctioning, jamming (Leachman 1997). In semiconductor industry, SEMI E10 (SEMI E10-0701 2001) is the guideline to specify the definition and measurement of equipment reliability, availability, and maintainability (RAM) for equipment performance. Figure 1 shows six major equipment statuses for measuring equipment RAM performance. The basic equipment statuses, including non-scheduled time, unscheduled downtime, scheduled downtime, engineering time, standby time, and productive time, could be further divided into as many sub-statuses as required to achieve the equipment tracking resolution for improving manufacturing operation objectives. Therefore, key blocks of time may associate with the basic statuses and sub-statuses particularly. SEMI E10 has been widely accepted as a set of industry-wide standards for buyers, suppliers, and manufacturers of semiconductor manufacturing equipment. By definition, OEE can be measured as follows: QE  RE OE  AE  timetotal unitseffective for time productionl theoretica OEE    )(  (1) where timetotal uptimeequipment  Efficiencyty Availabili  (AE)  ; uptimeequipment time production Efficiencyl Operationa  (OE)  ;  This paper has been published in  International Journal of Production Research . Reference to this paper should be made as follows: Chien, C.-F., Chen, H.-K., W, J.-Z., Hu, C.-H. (2007), “Constructing the OGE for promoting tool group productivity in semiconductor manufacturing,”  International Journal of Production Research , 45(3):509-524. (http://dx.doi.org/10.1080/00207540600792515) 4 time productionunitsactual  for time productionl theoretica  Efficiency Rate  (RE)  ; unitsactual  for time productionl theoretica unitseffective for time productionl theoretica  EfficiencyQuality  (QE)  . (Please insert Figure 1 about here) Furthermore, SEMI E79 defines OEE as the fraction of total time that equipment is producing effective units at theoretically efficient rates (SEMI E79-0200 2000). In particular, effective units denote the number of units processed by the equipment during production time that were of acceptable quality, i.e., actual unit output minus equipment assignable rework and scrap. Leachman (1997) defines rate efficiency as the ratio of the theoretical time to complete the work and the measured production time. As shown in Figures 2 and 3, the relationship between SEMI E10 and OEE can be structured in a hierarchical way. The E10 equipment statuses are classified into each efficiency component of OEE index. Each status also consists of several sub-statuses. The way of defining and classifying these sub-statuses varies by different users or equipment depending on the level of details, accuracy, and data collection ability. (Please insert Figure 2 and Figure 3 about here) While OEE has been widely accepted in semiconductor manufacturing industry, a number of related studies were conducted. For example, the srcinal OEE proposed by Nakajima (1998) did not consider the time scheduled for the preventive maintenance and the scheduled time to stop the plant. Thus, Jeong and Phillips (2001) proposed a loss classification scheme based on SEMI E10-92 to consider the time scheduled for the preventive maintenance and the scheduled time to stop the plant in capital-intensive industry. They also proposed a new interpretation for OEE including state analysis, relative loss analysis, lost unit analysis, and product unit analysis and then developed a methodology for constructing a data collection system for the total productivity improvement visibility system. Owing to considering the nonscheduled time and scheduled maintenance time, overestimation of OEE can be avoided to improve the OEE accuracy for capital-intensive industry such as semiconductor industry (Jeong and Phillips, 2001).  This paper has been published in  International Journal of Production Research . Reference to this paper should be made as follows: Chien, C.-F., Chen, H.-K., W, J.-Z., Hu, C.-H. (2007), “Constructing the OGE for promoting tool group productivity in semiconductor manufacturing,”  International Journal of Production Research , 45(3):509-524. (http://dx.doi.org/10.1080/00207540600792515) 5In addition, Oechsner et al. (2003) proposed the evaluation of Overall Factory Effectiveness (OFE) that considers different factors including the relationships between different machines and processes, integration of information, decisions, and actions across many independent systems and sub-systems. However, the complexity of semiconductor fabrication makes it difficult to employ OFE for evaluating overall fab effectiveness. Alternatively, Huang et al. (2003) proposed the effectiveness metrics of overall equipment effectiveness and Overall Throughput Effectiveness (OTE) for calculating equipment and system productivity for complex-connected manufacturing systems. Moreover, de Ron and Rooda (2005) proposed a new metric to compare and to improve tool productivity by excluding environment factors such as operator, recipe, facilities, material availability, scheduling requirements and used only effective time as the time base. 3. Proposed Overall Tool Group Efficiency (OGE) Although OEE has been increasingly applied in semiconductor manufacturing for managing performance of key tools, there is a gap between the efficiency measurement for individual equipment versus the overall efficiency of processes or the whole factory. Thus, focusing only on OEE improvement for individual machine may lead to undesired outcomes such as high WIP level and long cycle time. In addition, there are hundreds of machines in the factory and it requires much time and work to review each individual machine. Moreover, dispatching rules have been the most common shop-floor control tools in semiconductor manufacturing compared to complex approaches such as scheduling and control theories and knowledge-based systems (Uzsoy et al. 1994). Indeed, jobs are dispatched to the machine in a tool group while the other machines belonging to the same tool group may back up each other. Thus, low OEE of one machine may be caused by insufficient WIP, poor dispatching, or unbalanced loading among the machines in the same group. While considering tool productivities, multiple dispatching criteria should be combined to simultaneously achieve on-time delivery, low variance of lateness and variance of cycle time, and short mean cycle time (Dabbas et al. 2003). The difficulty can be traced at least in part to the lack of a modeling framework within which different interrelationships can be identified and related to the efficiency measurement for individual equipment and the overall efficiency of processes or the whole fab. To fill the gap, this study proposed a pair of performance indices collectively called Overall
Search
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks