Different manufacturing systems share some common approaches for dealing with changes and uncertainty, i.e., AUTONOMY, DISTRIBUTION, MODULARITY, and INTEGRATION. In this research, we analyze the relationships between these approaches. We used Structural Equation Modeling (SEM) to analyze sample data from various industries. The results show that AUTONOMY, DISTRIBUTION, and INTEGRATION support each other, i.e., a higher level in one approach can lead to a higher level in another approach. When we broke down the detail context of the MODULARITY, we found that different kinds of MODULARITY have different relationships with other approaches. Based on the results, we provided some suggestions for dealing with changes and uncertainty.

View full abstract

The proposal of the Equalized Lot Size Value Algorithm is introduced as an alternative to generate a better total makespan of a production system and, at the same time, pursuing the smoothing of the machine loading when a high demand fluctuation occurs. The ELSV algorithm combines the merits of the Equalized and Synchronized Production (ESP) and Theory of Constraints (TOC), in order to obtain a concrete definition of ESP algorithm and a smaller makespan for the Job Shop problems under TOC scheduling. As a result, smaller equalized lot sizes can improve makespan, especially for larger systems. On the other hand, equalized lot sizes also provide enough improvement in the smoothing of loading, even though it is a half of original size.

View full abstract

We present an extended model of JSSP that treats various heterogeneous decision features such as lot splitting (interruption of operation), dispatching as well as sequencing simultaneously. Consequently, the problem has a feature of combinatorial and continuous dynamic optimization and hence the extension of solution concept and a new solution principle are obligatory. One of the key ideas approaching to this problem is to introduce the concept of progress with regard to operation and to formulate it into a transient state in a dynamic optimization problem.

View full abstract

From the standpoint of operations management, we focus on a management cycle that consists of plan (P), do (D), check (C) and act (A) phases, and concern with the difference between PDCA and CAPD cycles. This paper studies a mathematical/stochastic modeling of the management cycle under lot processing and time span. This modeling is called the type of (t, k; T)-switching policy, in which the notations t, k and T means the review period, control limit and time span (due time), respectively. First, the switching modeling of the look-ahead (back) in PDCA (CAPD) is presented by stochastic approach, and corresponds to the trade-off problem of earliness (c_2) and tardiness (c_3) from time span (T). Next, the expected operating cost per unit produced is given by mathematical formulation, and the exponential processing type is treated as a special case of general service. Finally, the problem of optimal look-ahead/back strategy is numerically discussed and clarified on the base of(c_3, c_2)-strategy map in the fixed or variable T. In addition, an application to MRP/APS etc. is noted.

View full abstract

We have contrived these control charts in order to detect the out-of-control state of the process. The assignable causes of the out-of-control state consist of the incompleteness of materials supplied from the previous process or in the process operation at the present process. In quality management, it is important to detect the out-of-control state and to remove the factor by which it is caused. Although conventional control charts, in which the state of the process is judged only on output data of products from the present process, are useful statistical tools for detecting the out-of-control state, the procedure for identifying the factor that causes the out-of-control state has not been actively contrived. By the way, because there is a type II error in monitoring the previous process, it is clear that the incompleteness of materials can not be excluded from the assignable cause of the out-of-control state in the present process. Therefore, in this study, we consider the active use of the input data of materials provided in the present process in addition to the output data of products obtained from the present process, in order to both improve the detection of the out-of-control state of the present process and specify the factor of the assignable cause. First, we propose the new statistical procedure for monitoring the process state by using both the output data from the present process and the input data into the present process. Next, when the out-of-control state of the present process is detected by the proposed monitoring procedure, we consider the procedure for specifying the factor of the assignable cause of the out-of-control state. We also verify the effectiveness of the proposed monitoring system with both the judgment of the process state and the recognition of the assignable cause through numerical simulations. Further, we consider the economical operation of the proposed monitoring system based on the loss of the quality.

View full abstract

This paper deals with investment decisions under uncertainty and competition. Firms should make decisions about investment in order to maximize their expected payoff in the market. Then, their expected payoff is effected variously by the market condition where they are active. We especially consider uncertainty and competition in this paper. For such investment decision problems, the works that combine the real options approach and game theory are prevalent recently. Furthermore, investment decisions should have various stages with respect to the project life cycle. For example, the stages should have entry, expansion-reduction, stop-restart and exit. Existing literature have analyzed sufficiently entry stage. Therefore, we analyze optimal decision under uncertainty and competition by considering exit stage. In particular, we extend payoff functions in Huisman to exit stage. While rent equalization derives equilibria in Huisman, equilibria hold in spite of no rent equalization in this paper. Furthermore, the idea of results can be derived similarly to existing literature. In particular, the fact that the expected payoff of each firm is equal to that of follower, and that uncertainty delays decision. The new idea discovered in this paper is the fact that while competition hastens decision, it cannot result in the disappearance of the effect of uncertainty.

View full abstract

In this paper, we discuss a two-machine flowshop problem with the objective of minimizing the total flow time subject to minimum makespan. The problem is known as one of the hierarchical multicriteria scheduling problems which are a variant of the multicriteria scheduling problems. The problem is known to be NP-hard in general. T'kindt et al. provided a branch-and-bound algorithm to find an optimal schedule for the problem and reported that their algorithm solved problem instances effectively with up to 23 jobs. The purpose of this paper is to improve the branch-and-bound algorithm based on a generalization of a subproblem representation and new dominance relations. In most branch-and-bound algorithms for scheduling problems, a subproblem representation which fixes some jobs at the head of the schedule is used. In this paper, we use a more flexible subproblem representation in which jobs both at the head and at the end of the schedule are fixed. We show that existing dominance relations can be generalized and then propose some new relations. Dominance tests based on the dominance relations are incorporated into the branch-and-bound algorithm. The computational result shows that our algorithm works successfully to solve problem instances with up to 29 jobs. Furthermore, our algorithm outperforms an algorithm proposed by T'kindt et al. and a recent mixed integer programming solver.

View full abstract

In this paper, we propose a new idle time shortening method in a genetic algorithm (GA) for job-shop scheduling problems (JSPs). The purpose of JSP is to find a schedule with a minimum makespan. We suppose that it is effective to reduce idle time of a machine in order to improve the makespan. The shift left is the most famous method in existing methods for shortening idle time. The shift left arranges the work to idle time, without shifting another work to the right. For that reason the shift left can not arrange the work to idle time if it is shorter than the processing time of the work. We focused on such idle time. We propose an algorithm that expands the coverage of the shift left. We call this proposed algorithm the Eshift. The Eshift arranges the work to idle time that is shorter than the processing time. We propose two kinds of algorithms (g1 and g2) which generate a new chromosome from a new Gantt chart which is generated by the Eshift. These algorithms have the different features. For that reason, we combine these proposed algorithms. We put these proposed algorithms (called Pshift) into GA, and we test the validity of our Pshift by experiments using benchmark problems.

View full abstract

In this paper, we propose a method to translate cooking works into cooking robot recipe for a cooking robot system which consists of industrial robots, some sensors and computers. Content of cooking is summarized into table of cooking content by five items: cooking name, ingredient, cooking process, tool and equipment and processed ingredient after cooking and cooking condition of each item, which are collected from content of recipe and human's cooking knowledge. In this table, cooking process is classified into three cooking types based on aim of cooking work: preparation work, main work and arrangement work. Each cooking type is divided into robot cooking works based on aim of motion, that is, transportation of ingredients or tools, changing the shape or condition of ingredients and handling of equipment. Further, robot cooking work is divided into cooking motions consisted of robot motion and sensor motion. The validity of this robot recipe is evaluated by an experiment with our cooking robot system.

View full abstract