A simulation analysis for evaluating TFT-LCD fab capacity expansion with a distant transportation problem

Capacity planning for large-scale high-tech manufacturing processes such as semiconductor manufacturing and thin film transistor-liquid crystal display (TFT-LCD) using simulation of an entire fabrication facility (fab) requires a large computational effort and thus few studies have been in real settings. To address the needs of a realistic problem, this study aimed to develop an effective approach based on a discrete-event simulation model for evaluating the throughput, cycle time and utilisation in an integrated fab to integrate manufacturing and transportation resources. In particular, we conducted an empirical study in a real TFT-LCD fab expansion facing a difficult capacity planning problem arising from the expectation that one or more bottlenecks may shift to different sites, including the transportation system between the incumbent and the expansion fabs. Different product-mix alternatives and feeding policies are investigated to determine the best fab configuration. The results have shown practical viability of the proposed simulation technique to significantly reduce the computational effort associated with the capacity planning process and derive useful alternatives for supporting capacity expansion decisions.     

AMHS capacity determination model for wafer fabrication based on production performance optimization

Automatic material handling system (AMHS) is becoming more important in 300?mm wafer fabrication factories (fab). Effective and efficient design and control of AMHS has become more critical particularly in capacity planning. The major concept of the AMHS capacity determination model is to maintain the originally designed optimal production throughput or cycle time of products. In order to maintain fab’s throughput or cycle time of products, WIP (work in process) portfolio of the constraint or the fastest workstation should be kept. Based on this concept, a GI/G/m queuing model based on FCFS (First-come-first-serve) dispatching rule of AMHS is applied to determine the required number of vehicles. Basically, products should be transported to the specific workstation (constraint or fastest workstation) before the workstation finishes the existing process; therefore, sufficient WIP in front of this specific workstation should be kept. Under this condition, the probability that transportation time exceeds product processing time under a certain transportation capacity level can be calculated by the proposed model. Hence, we can get the required capacity of AMHS to achieve the probability target set in advance. Due to the capacity of AMHS can be set according to the acceptable probability of non-exceeding the processing time of the constraint or fastest workstation, the level of WIP in front of this workstation can be kept. It also can be ensured that AMHS will not affect the production performance as well as keep the reasonable investment level.     

Route planning for two wafer fabs with capacity-sharing mechanisms

This paper formulates and solves a route planning problem for semiconductor manufacturing. In order to quickly respond to rising demand, a semiconductor company usually adopts a dual-fab strategy to expand capacity. That is, two fab sites are built as neighbours and can easily share capacity. Through the capacity-sharing design, a product may be produced by a cross-fab route. That is, some operations of a product are carried out in one fab and the other operations in the other fab. This leads to a routing planning problem, which involves two decisions–determining the cut-off point of the cross-fab route and the route ratio for each product–in order to maximise the throughput subject to a cycle time constraint. An LP-GA method is proposed to solve the route planning problem. We first use the LP module to make the cut-off point decisions, and proceed to use the GA module for making the decision on the route ratio. Experimental results show that the LP-GA method significantly outperforms other methods.     

The simulation of air recirculation and fire/explosion phenomena within a semiconductor factory

The semiconductor industry is the collection of capital-intensive firms that employ a variety of hazardous chemicals and engage in the design and fabrication of semiconductor devices. Owing to its processing characteristics, the fully confined structure of the fabrication area (fab) and the vertical airflow ventilation design restrict the applications of traditional consequence analysis techniques that are commonly used in other industries. The adverse situation also limits the advancement of a fire/explosion prevention design for the industry. In this research, a realistic model of a semiconductor factory with a fab, sub-fabrication area, supply air plenum, and return air plenum structures was constructed and the computational fluid dynamics algorithm was employed to simulate the possible fire/explosion range and its severity. The semiconductor factory has fan module units with high efficiency particulate air filters that can keep the airflow uniform within the cleanroom. This condition was modeled by 25 fans, three layers of porous ceiling, and one layer of porous floor. The obtained results predicted very well the real airflow pattern in the semiconductor factory. Different released gases, leak locations, and leak rates were applied to investigate their influence on the hazard range and severity. Common mitigation measures such as a water spray system and a pressure relief panel were also provided to study their potential effectiveness to relieve thermal radiation and overpressure hazards within a fab. The semiconductor industry can use this simulation procedure as a reference on how to implement a consequence analysis for a flammable gas release accident within an air recirculation cleanroom.     

Estimating the sustainable capacity of semiconductor fab workstations

The cost of a semiconductor wafer fabrication facility (i.e. ‘fab’) may be as much as five or more billion US dollars. As such it is essential to determine the capacity (e.g. in terms of ‘wafer starts per week’) of such facilities. An accurate estimate of capacity – under real world conditions – is, however, difficult to achieve. Furthermore, the method for the computation of the capacity of a semiconductor fab is significantly different from that for the capacity of workstations in more conventional, less complex factories. This is due in part to the reentrant nature of the workstations (a.k.a. ‘toolsets’) that comprise a fab's production line as well as the ubiquitous employment of operation-to-machine dedications (a.k.a. layer-to-tool qualifications) – plus the need to consider multiple products employing, perhaps, a different sequence of processing steps. In this article, the matter of workstation capacity, in general, and semiconductor fabs, in particular, is examined. A means to quickly and effectively determine the maximum theoretical capacity of a workstation is developed and illustrated – followed by a way in which the more practical maximum sustainable capacity may be estimated.     

Scheduling semiconductor in-line steppers in new product/process introduction scenarios

This paper presents a scheduling method for an in-line stepper operating in a new process/production introduction (NPI) scenario. An in-line stepper is a bottleneck machine in a semiconductor fab. Its interior is comprised of a series of chambers, while its exterior is a dock equipped with a limited number of ports. The transportation unit for each chamber is a piece of wafer, while that for each port is a job that can contain up to 25 wafers. This transportation incompatibility may lead to an unexpected capacity loss for an in-line stepper–in particular in an NPI scenario that, by nature, includes a substantial number of small-sized jobs. Such a capacity loss can be alleviated by effective scheduling. A genetic algorithm (GA) scheduling method is proposed to enhance the productivity of in-line steppers. Four other sequencing methods are compared with the GA method. Numeric experiments indicate that the GA method outperforms the four benchmarks. The higher the percentage of small-sized jobs, the better the performance of the GA.     

A non-linear traffic flow-based queuing model to estimate container terminal throughput with AGVs

Efficient handling of containers at a terminal can reduce the overall vessel sojourn times and minimise operational costs. The internal transport of containers in these terminals is performed by vehicles that share a common guide path. The throughput capacity of a terminal may increase by increasing the number of vehicles; however, simultaneously congestion may reduce the effective vehicle speed. We model this situation accurately using a traffic flow-based closed queuing network model. The vehicle internal transport is modelled using a load-dependent server that captures the interaction between the number of vehicles in a transport segment and the effective vehicle speed. Using a non-linear traffic flow model, we show that the throughput reductions due to vehicle congestion can be as large as 85%. Hence, the effect of vehicle congestion during internal transport cannot be ignored. The model can also be used to determine the appropriate number of vehicles required to achieve the required terminal throughput by explicitly considering the effect of vehicle congestion.     

Determining inventory levels in a CONWIP controlled job shop

We extend the concept of CONWIP control to a job shop setting, in which multiple products with distinct routings compete for the same set of resources. The problem is to determine the fixed overall WIP level and its allocation to product types (WIP mix) to meet a uniformly high customer service requirement for each product type. We formulate an optimization problem for an open queuing network model in which customer orders pull completed products from the system. Then, assuming heavy demand, we derive a throughput target for each product type in a closed queuing network and provide a simple heuristic to find a minimum total WJP and WIP mix that will achieve an operating throughput close to this target. In numerical examples, the WIP mix suggested by this approach achieves the customer service requirement with a relatively low total WIP     

A multi-criteria decision-making approach for capacity allocation problem in semiconductor fabrication

In today's highly competitive environment, good customer relationship management is essential for a company to survive and to acquire reasonable profit. A semiconductor enterprise has multi-site fabs, and follows basically a make-to-order production type. The allocation of capacity for manufacturing different types of products is important for the competitiveness and future development of the enterprise. To cope with this requirement, a multi-criteria decision-making approach is proposed in this article for more efficient evaluation and selection of capacity allocation plans in semiconductor fabs. Under this approach, fuzzy analytic network process (FANP) is incorporated with fuzzy Delphi method (FDM), constraint programming (CP) and benefits, opportunities, costs and risks (BOCR). FDM is used to achieve a consensus among the opinions of experts. CP is used to screen out a set of capacity allocation plans that should be further evaluated. FANP with BOCR is used to evaluate the various factors and the interrelationship among the factors under the BOCR merits and to determine the expected performance of the capacity allocation plans. The proposed approach can provide a ranking and also priorities of different capacity allocation plans, and the fab can select the most appropriate capacity allocation for production in the case that insufficient capacity existed.     

Modelling and design for a shuttle-based storage and retrieval system

The shuttle-based storage and retrieval system (SBS/RS) is a relatively new part-to-picker order picking system. We have developed a performance estimation and design algorithm for the SBS/RS. The performance estimation is based on a queuing model. The design algorithm aims to find the minimum cost configurations in terms of number of tiers, aisles, lifts and workstations with given throughput, tote capacity and order cycle time requirements. We used simulation driven by parameters abstracted from an actual SBS/RS to verify the performance estimation, and applied the design algorithm in the case study. The results indicate that: (1) compared to simulation results, the throughput of the performance estimation is nearly identical when the arrival rate is below the maximum capacity; (2) the design algorithm yields a configuration with 28.1% cost reduction in the current system. In addition, we also compared the shuttle system with the competing robotic order fulfilment system (robotic system in short) in terms of facility cost, building cost and order cycle time. We found that the shuttle system is a better choice if large storage capacity and high throughput are required whereas the robotic order fulfilment system performs better if small storage capacity and low throughput are required.     

Optimal stack layout in a sea container terminal with automated lifting vehicles

Container terminal performance is largely determined by its design decisions, which include the number and type of quay cranes, stack cranes, transport vehicles, vehicle travel path and stack layout. We investigate the orientation of the stack layout (parallel or perpendicular to the quayside) on the throughput time performance of the terminals. Previous studies in this area typically use deterministic optimisation, and a few studies use probabilistic travel times and simulation to analyse the effect of stack layout on terminal throughput times. In this research, we capture the stochasticity with an integrated queuing network modelling approach to analyse the performance of container terminals with parallel stack layout using automated lifting vehicles. Using this analytical model, we investigate 1008 parallel stack layout configurations on throughput times and determine the optimal stack layout configuration. We find that, assuming an identical width of the internal transport area, container terminals with parallel stack layout perform better (from 4–12% in terms of container throughput times) than terminals with a perpendicular stack layout.     

Incorporating engineering process improvement activities into production planning formulations using a large-scale wafer fab model

In most semiconductor wafer fabrication facilities (wafer fabs), both production and engineering lots share the same expensive equipment. Production lots will be shipped to customers whereas engineering lots are used to support production and process development efforts. While production activities might lead to large revenue, engineering activities result in increased future output. In the present paper, we propose three different production planning formulations. The first formulation assumes a reduced available capacity for production due to engineering activities. Costs for production products are minimised. The second formulation is based on the idea that aggregated demand is available for engineering activities for the entire planning window. Costs for production and engineering products are minimised. Learning effects are incorporated to model the capacity increase due to engineering activities. The third model assumes that demand information for engineering activities is available only for the first period. In addition to learning effects, forgetting effects are modelled to provide an incentive to plan releases of engineering lots in later periods. Costs for production and engineering products and forgetting effects are minimised. The performance of the production planning models is assessed using a simulation model of a large-scale wafer fab including specific dispatching strategies to deal with production and engineering lots. The simulation results demonstrate that the second model slightly outperforms the third one when a rolling horizon approach is taken, while the second model provides significantly higher profit than the first one.     

Queuing network analysis approach for estimating the sizes of the time buffers in Theory of Constraints-controlled production systems

The present paper describes an open queuing network modelling approach to estimate the size of the time buffers in production systems controlled by the Theory of Constraints philosophy. Workstations in the production network are modelled as GI/G/m queues and a queuing network analysis multiproduct open queuing network modelling method is used to estimate the average flow time to the time buffer origin and the standard deviation of flow time. Using these two values together with an assumption of normally distributed flow times and a chosen service level, the final time buffer length is determined. The queuing network analysis method has been modified to enable the modelling of production networks with machine failures, batch service and varying transfer batch sizes. The modelling approach has also been incorporated in a computerized tool that uses product specific information such as bill-of-material and routing data, and production network information such as resource data to estimate the sizes and location of the necessary time buffers for each product. Simulation experiments indicate that the procedure is sufficiently accurate to provide an initial quick estimate of the needed time buffer lengths at the design stage of the line.     

Coupling analysis of passenger and train flows for a large-scale urban rail transit system

Coupling analysis of passenger and train flows is an important approach in evaluating and optimizing the operation efficiency of large-scale urban rail transit (URT) systems. This study proposes a passenger–train interaction simulation approach to determine the coupling relationship between passenger and train flows. On the bases of time-varying origin–destination demand, train timetable, and network topology, the proposed approach can restore passenger behaviors in URT systems. Upstream priority, queuing process with first-in-first-serve principle, and capacity constraints are considered in the proposed simulation mechanism. This approach can also obtain each passenger’s complete travel chain, which can be used to analyze (including but not limited to) various indicators discussed in this research to effectively support train schedule optimization and capacity evaluation for urban rail managers. Lastly, the proposed model and its potential application are demonstrated via numerical experiments using real-world data from the Beijing URT system (i.e., rail network with the world’s highest passenger ridership).     

机场容量限制下的多重分派枢纽航线网络设计

为了探索枢纽航线网络的设计,提出了机场容量限制下的多重分派枢纽航线网络的设计方法。以航空客运O-D流为基础,将枢纽选择和航线设计两个阶段综合考虑,用机场吞吐量衡量机场容量限制,建立多商品网络流模型,量化成本折扣因子、枢纽数目对枢纽航线网络总成本的影响程度。中国航空客运吞吐量排名前20的城市组成算例,用IBM ILOG CPLEX12.2计算出设计方案的时间不超过14.87 s。算例证明,机场容量限制会增加枢纽航线网络的总成本,但O-D流分布过度密集的现象减轻,抗拥堵性增强。枢纽间成本折扣因子的减小和枢纽数目的适当增加都能够降低枢纽航线网络的总成本。最后,绘制枢纽航线网络结构图,为枢纽航线网络的设计提供指导。     

面向定制生产系统的缓冲区优化配置方法

针对定制型装备制造企业中具有有限缓冲区的开排队网制造单元,其车间负荷界限即缓冲设置难以确定的问题（buffer allocation problem,BAP）,文章对每阶段具有有限缓冲区且含有多台加工设备的三阶段柔性流水车间(flexible flow shop,FFS)进行排队网建模,应用状态空间分解法对该模型进行分析求解,获得系统的一系列性能指标值。为了对该方法的有效性进行验证,对该模型设计仿真实验,并利用扩展法对模型进行求解,将数值结果进行比较分析,验证了利用该方法对FFS缓冲区进行优化配置的合理性,这对较大规模的多节点每阶段具有多台设备的流水车间负荷界限的有效设定及其规划具有参考和指导意义。     

A practical fab design procedure for wafer fabrication plants

Facility design is crucial to the performance of wafer fabs in the semiconductor industry. This research proposes a practical Fab Design Procedure (FDP) to conduct quick calculations to develop and evaluate initial fab design alternatives. A series of practical formulae are presented to sequentially determine the following design parameters from an interbay point of view: the required number of machines; machine grouping and allocation; interbay flow matrix; bay dimension and location; interbay distance matrix; the required number of vehicles in an interbay material handling system; and the average wafer moving distance. Rules-of-thumb for wafer fab design on the basis of FDP are also suggested. A case study is used to demonstrate the effectiveness and efficiency of FDP. Results indicate that FDP is able to quickly calculate the required number of machines and related fab design parameters. With this tool, fab designers would be able to evaluate design alternatives and conduct what-if analysis in the initial phase of fab design.     

Optimizing the intermodal transportation of emergency medical supplies using balanced fuzzy clustering

In this paper, we are concerned with the problem of the ‘helicopters and vehicles’ intermodal transportation of medical supplies in response to large-scale disasters. To deal with the disadvantages of the use of classic Fuzzy C-Means (FCM) in the intermodal transportation optimization, two balanced FCM methods, i.e. FCM with capacity constraints and FCM with number constraints, are formulated to select emergency distribution centers (EDCs) and assign medical aid points, which could construct balanced ‘helicopters and vehicles’ intermodal transportation network. Then, considering helicopter travel time, transfer time and vehicle delivery time, a clustering-based intermodal routes optimization model is presented to produce intermodal transportation routes. Numerical experiments are presented to show the effectiveness and advantage of the developed approach, and observe the impact of number of EDCs and transfer efficiency at EDCs on the performance of intermodal transportation. This paper could provide methodological and operational supports for the ‘helicopters and vehicles’ intermodal transportation of medical supplies in response to large-scale disasters.     

航线运输驾驶员转场训练网络优化设计

为了改进传统航线运输驾驶员转场训练网络设计方法的不足之处,以降低转场训练网络的设计总成本,构建了包含源发弧、始发弧、衔接弧和到达弧在内的转场训练路径衔接网络,并利用计算机深度遍历算法获取各类转场训练科目的可行训练路径集合,然后以训练基地选址和流经各转场训练路径上的航线驾驶员培训数量为决策变量,并综合考虑训练基地设置数量、航线运输驾驶员培训总人数和各训练基地转场训练容量等限制条件,以转场训练网络总成本最小为目标函数,构建转场训练网络优化设计数学模型。算例结果表明,与传统方法相比,该方法将转场训练网络总成本降低了9.18%;通过将空域及机场容量在−30% ~ 30%之间波动,发现转场训练网络中训练基地数量发生变化,并使得网络设计总成本在1.12% ~ −10.67%之间波动。因此,该方法是可行且有效的,能够为飞行航校的转场训练网络设计提供指导。     

Semiconductor capacity planning: stochastic modelingand computational studies

This paper presents a multistage stochastic programming model for strategic capacity planning at a major US semiconductor manufacturer. Main sources of uncertainty in this multi-year planning problem include demand of different technologies and capacity estimations for each fabrication (fab) facility. We test the model using real-world scenarios requiring the determination of capacity planning for 29 technology categories among five fab facilities. The objective of the model is to minimize the gaps between product demands and the capacity allocated to the technology specified by each product. We consider two different scenario-analysis constructs: first, an independent scenario structure where we assume no prior information and the model systematically enumerates possible states in each period. The states from one period to another are independent from each other. Second, we consider an arbitrary scenario construct, which allows the planner to sample/evaluate arbitrary multi-period scenarios that captures the dependency between periods. In both cases, a scenario is defined as a multi-period path from the root to a leaf in the scenario tree. We conduct intensive computational experiments on these models using real data supplied by the semiconductor manufacturer. The purpose of our experiments is two-fold: first to examine different degree of scenario aggregation and its effects on the independent model to achieve high-quality solution. Using this as a benchmark, we then compare the results from the arbitrary model and illustrate the different uses of the two scenario constructs. We show that the independent model allows a varying degree of scenario aggregation without significant prior information, while the arbitrary model allows planners to play out specific scenarios given prior information.