Search

For those products that are heavily competitive in the marketplace, demand volatility and unpredictability have been growing. This has resulted in a sizeable deviation in demand forecasts when using the traditional forecasting methods. Accordingly, this study aims to develop a real option approach-based forecasting model for predicting demand during the upcoming planning horizon for products with high random volatility on demand. The real option approach can effectively deal with the long-term trends and random variation involved in a given demand stochastic diffusion process. Additionally, this study proposes taking Monte Carlo simulation as a numerical method to solve the demand-forecasting model. Monte Carlo simulation not only can accurately approximate almost any type of stochastic processes, but also can competently handle the path-dependant relationship existing between successive demands. Subsequently, these demand forecasts are used to determine the provisioned smoothing capacity during the upcom...

With demand variability and unpredictability, decreasing product life cycles, globalisation and increasing competition, rush orders are increasingly being received at job shops. However, these rush orders can be difficult to fulfil owing to a lack of capacity during boom and heavy workload periods. In this situation, a feasible solution is capacity exchange, through which capacity is transacted and exchanged with other customers for whom the manufacturer has already reserved capacity. Accordingly, this study develops a customer-driven capacity exchange mechanism for solving the foregoing problem of temporary capacity shortage caused by the receipt of rush orders in a job shop. Notably, this innovative concept and mechanism can be expected to concurrently benefit capacity holders, capacity demanders and manufacturers. Additionally, this study presents a comprehensive design scheme, working scenario, and possible implementation for a capacity exchange mechanism.

The complicated nature of the semiconductor manufacturing process requires high precision and thousands of spare parts. Due to the characteristics of high cost, as well as dynamic demand rate for spare parts of semiconductor equipment, spare parts inventory management is always an important issue for the equipment supplier. In this paper, we focus on the case of spare parts with high cost and low usage. Using varying lead times and machine life cycles, a simple model to evaluate the relationship between inventory level and service level is developed and a moving sum method is used to estimate the parameters of the model while only a small set of demand data is available. Then we take Applied Materials Taiwan (AMT), a major semiconductor equipment manufacturer in Taiwan, as an example to demonstrate the estimation of parameters of lead time distribution and parts demand rate and the implementation of the inventory model.

This study investigates the buffer allocation strategy of a flow-shop-type production system that possesses a given total amount of buffers and finite buffer capacity for each workstation as well as general interarrival and service times in order to optimize such system performances as minimizing work-in-process, cycle time and blocking probability, maximizing throughput, or their combinations. In theory, the buffer allocation problem is in itself a difficult NP-hard combinatorial optimization problem, it is made even more difficult by the fact that the objective function is not obtainable in closed form for interrelating the integer decision variables (i.e., buffer sizes) and the performance measures of the system. Therefore, the purpose of this paper is to present an effective design methodology for buffer allocation in the production system. Our design methodology uses a dynamic programming process along with the embedded approximate analytic procedure for computing system performance measures under a certain allocation strategy. Numerical experiments show that our design methodology can quickly and quite precisely seek out the optimal or sub-optimal allocation strategy for most production system patterns.

This study extends previous results for batch-service workstations to batch-service/batch-lot workstations with multiple process recipes, e.g., diffusion operations in semiconductor manufacturing. The model considered herein explicitly considers the existence of a manufacturing operation associated with multiple process recipes in the semiconductor factory. Consequently, the revised balance equations are submitted and an improved approximation is presented for this case. Based on a comparison with simulation results, this new approximation is shown to be superior to the previously developed analytical approaches. This new approximation is especially strong in cases where the number of process recipes grows, system traffic intensity is moderate, and arrival rate of each recipe is nearly the same.

To investigate the therapeutic effects of different skin flaps for repairing severe complex hand injuries with burns and compression.From January 1990 to December 2000, 39 patients with severe complex injuries due to burns and compression in the hand were treated with different skin flaps, followed by early-stage postoperative comprehensive rehabilitation therapy.All the patients were followed up for 6 to 12 months, and significant differences were observed in the appearance and function of the repaired hands with different skin flaps.The medial skin flap taken from the lower leg and the reverse-flow island flap from the anterior aspect of the forearm, when combined with early postoperative rehabilitation treatment, can achieve good recovery of function and appearance of the injured hand.

To evaluate the clinical effect of total hip joint replacement and hemiarthroplasty in treatment of fracture of femoral neck in old patients.One hundred and ten cases with femoral neck fracture in the aged, 70 to 106 years old, from Aug 1990 to Aug 1999 were reviewed, 96 cases were followed up, among which 52 cases received total hip joints replacement and the other 44 cases received hemiarthroplasty. All of the 96 cases were followed up for 15 to 112 months, averaged 51 months, and were evaluated in operation procedures, post-operative recovery and joint function according to Harris Scoring.The operation time of total hip joints replacement was 20 minutes longer, bleeding volume was 120 ml larger, and post-operative drainage was 140 ml more, in average, than those in hemiarthroplasty. There was no obvious difference between the two types of operation in bed-resting time, length of stay and hospitalizing costs. According to Harris Scoring, there were 38 cases of excellent in hemiarthroplasty (86.4%) and 48 cases of excellent in total hip joints replacement (92.3%).Both of the artificial joint replacements are reasonable choices for treatment of fracture of femoral neck in old patients, but total hip joints replacement is recommendable for those comparatively younger patients with good systematic status, and hemiarthroplasty is a good option for those elderly with some systematic diseases.