A project is always performed in an environment influenced by resource and environment factors such as equipment, weather, labor skill, materials delivery, management quality and so on. Since these factors affect various activities on the same path in an activity network, activity durations tend to vary together. The variation of these durations affect the path duration and/or the project duration. This paper introduces a new method that incorporates these factors in a fuzzy activity network and provides factor sensitivity information to support efficient project schedule management. This method uses a triangular fuzzy number for the activity duration, and uses the spread of the triangular fuzzy number for the variation in the estimation of the activity duration. Based on the qualitative estimations of the sensitivity of each activity to each factor, the variation in activity duration is distributed to several factor variations (f-variation). Each f-variation is then broken down into more variations based on factor conditions (c-variations), with each c-variation corresponding to a factor condition such as better-than-expected, expected, or worse-than-expected. The effects of resource and environment factors are captured using c-variations under the same factor conditions to compute path and project durations. The proposed method makes it possible to improve the estimation of the path and project durations. The sensitivity of paths to factors can be searched and effectively used to improve project schedule management. Finally, an example is demonstrated, and the results are compared with other methods.

View full abstract

A synchronized production system has been built that corresponds to a model virtualized on PLC memory with production processes for a modular design product. Retention of information as well as material could be eliminated by exchanging real-time information on production model information and production information, applying cycle time matching for processes, automatic setup change for processes, and inlining. We would like to introduce a case which has materialized higher production efficiency through a synchronized production system.

View full abstract

The configuration of a production system affects its performance. There has been much research on configuring production systems, such as assembly line balancing and cell formation. Under a specified configuration, the performance of the production system is affected by production ordering and buffer allocation. Ordering systems determine the order timing and materials quantity released for each stage of a multi-stage production inventory system and buffer volumes are often allocated; just-in-time (JIT) ordering systems determine them on the basis of actual demand. The Kanban system was developed by a Japanese automobile manufacturer as an original JIT ordering system. Alternatives include the constant work-in-process (CONWIP) system and the concurrent ordering system. The performance of JIT ordering systems is analyzed and compared with other systems. In most of the previous literature on JIT ordering systems, stable changes in demand have been assumed, and the influence of unstable changes in demand had never been analyzed. Based on this background, some reactive systems have been proposed for the Kanban and CONWIP systems. In reactive Kanban or CONWIP systems, unstable changes in demand are detected, and the number of Kanbans and the buffer size are adjusted as a reaction to the changes detected. However, if only the number of Kanbans or the buffer size is adjusted, adaptability is limited. Recently, reconfigurable manufacturing systems (RMSs) have been introduced, and systems which can adjust the configuration in response to environmental changes are likely to be realized. Therefore, this paper introduces dynamic line balancing and buffer allocation in a RMS controlled by the CONWIP system. In the proposed system, unstable changes in demand are detected by exponentially weighted moving average (EWMA) charts, the configuration of the production system-that is, the number of stages and buffer size-are altered in response to the unstable changes detected, and work elements are allocated to each stage by balancing the line. The effectiveness of the proposed system is analyzed and compared with that of the previous system using simulation experiments.

View full abstract

This paper discusses a dynamic lot size scheduling problem in which the set up operation assumes a sequence dependent aspect because each item simultaneously belongs to two distinct attribute families, yet for each attribute, set up operation is required as the families are switched. First, we present a modeling method for reducing the problem to a certain multi-item multi-process dynamic lot size scheduling model with sequence independent set up by relaxing some constraints and assigning other complementary constraints. Second, we present a numerical solution method using the Lagrangian decomposition coordination method together with a heuristics.

View full abstract

The widespread use of the online retail market in recent years has resulted in the share of online sales by e-tailers increasing in the areas of industrial/commercial business. When using online sales, some goods that are delivered look different from pictures/descriptions in the catalog. Therefore, online sales tend to have a much higher return rate than brick-and-mortar sales. In a very short selling period, the returned goods can be in "like new" condition, so returns may be reused to satisfy a new order as a secondary supply channel. But this is only possible if the goods are available before the end of the selling period. Vlachos & Dekker discussed return handling options under a high return rate and an approach to determine the optimal initial order quantity for a single period product when the e-tailer independently deals with return processing operations. In this case, however, the return procedures by the e-tailer may not be smooth because of poor handling of returns, and it may follow that the e-tailer's costs related to various return processing operations is extremely high. In order to smooth the return processing operations and reduce the cost related to the return processing that e-tailers incur, outsourcing service agencies that handle return processing operations are being increasingly used by e-tailers. In this paper, we discuss a return processing operation based on the e-tailer using an outsourcing service agency and develop analytical expressions of this case to determine the optimal order quantities. In order to assess the utility of outsourcing service agency that handles returns, we compare the optimal order quantities and the expected profit obtained from an e-tailer that uses an outsourcing service agency with those obtained from the e-tailer's own through results of numerical examples.

View full abstract

Nowadays, many Japanese companies introduce cellular manufacturing in their plants as new manufacturing systems in order to cope with the market competition for multi-item and small-batch products. At the same time, shorter product life cycles and increased demand for customization have reduced the viability of conveyer assembly line as a way to meet these requirements. These demands are expected to continue in the future, requiring systems that can quickly manufacture small batches of customized products in a cost-effective manner. Cellular manufacturing is considered as a new system to achieve these goals. Against the conveyer assembly line system in which a large number of workers are arranged at sequent workstations and the product moves from one station to the next until it is finished at the end of the line, cellular manufacturing needs only a small number of workers to carry out the assembly operation of a product continuously from start to finish. There are a lot of documented reports which described the advantages of cellular manufacturing, such as the viability of handling multi-item small-sized products, improvement in productivity, shortened lead time, saved work space, increase in worker motivation, and so on. However, converting assembly line to cellular manufacturing is not an easy matter because there are a lot of control factors to be considered. In fact, some companies are not able to accomplish performance improvements by converting their assembly lines to cells. The reasons for the improvement resulting from conversion have not been well documented or understood, making it difficult to know when and where cells are applicable. In this paper, we propose a theoretical study to analyze the main control factors of converting assembly line to cellular manufacturing. Most of the control factors are reported in previous documents but with only a very limited viewpoint. In our model, additional tasks, cross work training, work difficulty, the difficulty of increasing of tasks and the degree of performance of a work support system are discussed when converting assembly line to cellular manufacturing. Moreover, we also discuss productivity and the product inventory of each production method by the simulation that uses various product mixes. Assuming work task and worker's skill are probability variables, simulation models based on data collected from the previous documents are then used to estimate the marginal impact each factor change had on the estimated performance improvement resulting from the conversion to cells. The simulation results can clarify the relationships among these control factors and give a fundamental indicator of manufacturing system selection.

View full abstract

The vehicle routing problem (VRP) is one of the representative problems in mathematical programming, and it has various applications in many fields. Therefore, many studies have been made on VRP. In most of these studies, it is implicitly assumed that each vehicle is used for only one route. However, in practice, multiple use of vehicles is often allowed. Also, in most previous studies, the objective function is to minimize the total distance traveled by vehicles. However, there exist practical cases where it is important to minimize the length of the longest time traveled by any vehicle. In this paper, we consider a vehicle routing problem with multiple use of vehicles, where the objective function is to minimize the longest travel time. We propose a near-optimal solution of this problem on the basis of the Lagrangian heuristic. First, the problem is formulated as a mixed 0-1 integer programming problem, and a method to determine the lower bound of the problem is shown by solving a Lagrangian relaxation problem. Next, we propose a method to determine a near-optimal solution of the problem by successively decreasing the difference between upper and lower bounds (duality gap). Finally, we verify the effectiveness of the proposed method through computational experiments.

View full abstract

It is important to know whether or not performance ratings provide an accurate reflection of the performance of individuals being rated. Despite the accuracy of performance ratings, it is generally acknowledged that they are too often contaminated by systematic rater errors (i.e., leniency, central tendency, severity, etc.). Rater training is an area that has shown some promise in improving the effectiveness of performance ratings. The purpose of this study is to empirically examine the effectiveness of rater training measured by rating errors as psychometric outcomes in rating and the effect of reducing errors in the rater judgment process through the feedback of two things; (1) quantitative errors of each rater with factor scores that are estimated by a structural equation model created by combining two models; one being a latent curve model and the other a confirmatory factor analysis model, and (2) whether or not the rater committed errors. The results of the experiment attended by a total of 116 business persons clearly show that raters who committed rating errors the first time reduced the number of error the second time as the result of the above-mentioned feedback. This suggests two things; rater training itself is effective for reducing rating errors, and the training process is effective for reducing rater errors.

View full abstract

Recently, Toyota Production System is being introduced more application. As a multi-process holding system is used in Toyota Production System, work including standing and walking is needed. This paper studies these working conditions employing female high-school graduates as the objects of research. The period of the study was 16 weeks. The postures of experienced female workers and those of newly employed female workers were compared. The basic work task was sewing. In the study, the Questionnaire for Subjective Fatigue Feeling by Industrial Fatigue Research Committee of Japan Association of Industrial Health was used. As a result, it was found that they self-consciously felt fatigue in the first week, but it gradually disappeared. It increased again at Golden Week (consecutive holidays at the end of April and beginning of May), but subsided rapidly. Although there was significant fatigue reported for the feet and waist, it subsided with the passing of time similarly. The self-consciousness of fatigue from standing including walking was lower than that of standing, and no higher than that of sitting in a chair. Considering the above facts, it is believed that work while standing including walking causes no excessive burden from the aspect of self-conscious fatigue, and employees can adopt to such work conditions quickly.

View full abstract