Search results

Demand forecasting has long been an imperative tenet in production planning especially in a make-to-order environment where a typical manufacturer has to balance the issues of holding excessive safety stocks and experiencing possible stockout. Many studies provide pragmatic paradigms to generate demand forecasts (mainly based on smoothing forecasting models.) At the same time, artificial neural networks (ANNs) have been emerging as alternatives. In this chapter, we propose a two-stage forecasting approach, which combines the strengths of a neural network with a more conventional exponential smoothing model. In the first stage of this approach, a smoothing model estimates the series of demand forecasts. In the second stage, general regression neural network (GRNN) is applied to learn and then correct the errors of estimates. Our empirical study evaluates the use of different static and dynamic smoothing models and calibrates their synergies with GRNN. Various statistical tests are performed to compare the performances of the two-stage models (with error correction by neural network) and those of the original single-stage models (without error-correction by neural network). Comparisons with the single-stage GRNN are also included. Statistical results show that neural network correction leads to improvements to the forecasts made by all examined smoothing models and can outperform the single-stage GRNN in most cases. Relative performances at different levels of demand lumpiness are also examined.

This study examines the scheduling problem for a two-stage flowshop. All jobs are immediately available for processing and job characteristics including the processing times and due dates are known and certain. The goals of the scheduling problem are (1) to minimize the total flowtime for all jobs, (2) to minimize the total number of tardy jobs, and (3) to minimize both the total flowtime and the total number of tardy jobs simultaneously. Lower bound performances with respect to the total flowtime and the total number of tardy jobs are presented. Subsequently, this study identifies the special structure of schedules with minimum flowtime and minimum number of tardy jobs and develops three sets of heuristics which generate a Pareto set of bicriteria schedules. For each heuristic procedure, there are four options available for schedule generation. In addition, we provide enhancements to a variety of lower bounds with respect to flowtime and number of tardy jobs in a flowshop environment. Proofs and discussions to lower bound results are also included.

The primary purpose of this study is to illustrate a practical approach for industrial work process design that, in an integrative manner, captures essential concerns from different parties associated with manufacturing. It aims explicitly to incorporate utility expectation from the perspectives of operational managers, floor workers, and financial planners into the decision making process.

Through a real industrial scenario, the case study illustrates the use of a Bayesian belief network (BBN)‐based expert system and influence diagram in work process design. What‐if analysis is performed. Statistical tests are then used to benchmark and validate the experimental results and actual data.

The results suggest that the proposed BBN framework is effective in modeling and solving the work design problem. The findings can draw meaningful insights into the adoption and capacity of BBN in the fields of ergonomics, worker health management, and performance improvement.

Practically, the industrial problem is to compare the new stand‐up sewing cells against the traditional sit‐down sewing layout while taking into consideration of ergonomic effect (repetitive motion injury (RMI) likelihood), floor space (SF), yield (%), and cost ($). The study illustrates the use of an expert system and influence diagram to evaluate different alternatives for ergonomic work design in production process.

The results of this study can potentially improve health safety management and worker ergonomics.

The paper is among the few systematic studies that have applied BBN and influence diagram to production ergonomics and worker health management. A methodological framework utilizing these probabilistic reasoning techniques are developed. This new framework can capture essential concerns from different parties in manufacturing.

The purpose of this paper is to identify, study and quantify the effects of lighting on yield and productivity in manual electronics assembly (MEA) and inspection as a limiting work design criterion. The study also examines the potential interactions among lighting option, workers' age, and years of experience.

A three‐factor full factorial experiment is adopted to statistically evaluate the independent variables (process yield and assembly time) versus randomly selected levels of three factors: type of light (low pressure sodium, mercury vapor lamps, and metal halide lamps measured in foot‐candle luminaries), operator age, and years on the job. A residual analysis is also conducted to complement and corroborate the ANOVA findings.

The study finds that metal halide lamps, based on the ANSI recommended ranges of 186‐464 foot‐candles, lead to significant increases in labor productivity and through‐put, irrespective of operators' age and years of experience. Although these lamps have a significantly shorter life span than that of low‐pressure sodium and mercury vapor lamps, the realized benefits far exceed the incremental cost of illumination devices. The results indicate that a modest capital investment is able to generate solid improvements in yield and processing time in a typical MEA environment.

The relations between productivity and lighting intensity and type have never been studied in the area of MEA. This empirical study uncovers the effects through a systematic experimentation of this essential relationship in a typical MEA environment. The findings, which can be generalized to other facilities, are validated by an array of statistical procedures and proved to be significant. The paper contributes useful knowledge to the fields of engineering management and facility design.

Increasing competition within the global supply chain network has been pressuring managers to improve efficiencies of production systems while, at the same time, reduce manufacturing operation expenses. One well-known approach is to have better control of the manufacturing system through more accurate forecasting and efficient control. In other words, a production control paradigm with more reliable forward visibility should help in maintaining a cost-effective yet lean manufacturing environment. Hence, this study proposes a predictive decision support system for controlling and managing complex production environments and demonstrates a Visual Interactive Simulation (VIS) framework for forecasting system performances given a designated set of production control parameters. The VIS framework is applied to a real-world manufacturing system in which the primary objective is to minimize total production while maintaining consistently high throughput and controlling work-in-process level. Through this case study, we demonstrate the use and validate the effectiveness of VIS in optimization and prediction of the examined production system. Results show that the predictive VIS framework leads to better and more reliable decision making on selection of control parameters for the manufacturing system under study. Statistical analyses are incorporated to further strengthen the VIS decision-making process.

Most setup management techniques associated with electronic assembly operations focus on component similarity in grouping boards for batch processing. These process planning techniques often minimize setup times. On the contrary, grouping with respect to component geometry and frequency has been proved to further minimize assembly time. Thus, we propose the Placement Location Metric (PLM) algorithm to recognize and measure the similarity between printed circuit board (PCB) patterns. Grouping PCBs based on the geometric and frequency patterns of components in boards will form clusters of locations and, if these clusters are common between boards, similarity among layouts can be recognized. Hence, placement time will decrease if boards are grouped together with respect to the geometric similarity because the machine head will travel less. Given these notions, this study develops a new technique to group PCBs based on the essences of both component commonality and the PLM. The proposed pattern recognition method in conjunction with the Improved Group Setup (IGS) technique can be viewed as an extended enhancement to the existing Group Setup (GS) technique, which groups PCBs solely according to component similarity. Our analysis indicates that the IGS performs relatively well with respect to an array of existing setup management strategies. Experimental results also show that the IGS produces a better makespan than its counterparts over a low range of machine changeover times. These results are especially important to operations that need to manufacture quickly batches of relatively standardized products in moderate to larger volumes or in flexible cell environments. Moreover, the study provides justification to adopt different group management paradigms by electronic suppliers under a variety of processing conditions.