Heuristic optimization algorithms seek good feasible solutions to optimization problems in circumstances where the complexities of the problem or the limited time available for solution do not allow exact solution. Although worst case and probabilistic analysis of algorithms have produced insight on some classic models, most of the heuristics developed for large optimization problem must be evaluated empirically—by applying procedures to a collection of specific instances and comparing the observed solution quality and computational burden. This paper focuses on the methodological issues that must be confronted by researchers undertaking such experimental evaluations of heuristics, including experimental design, sources of test instances, measures of algorithmic performance, analysis of results, and presentation in papers and talks. The questions are difficult, and there are no clear right answers. We seek only to highlight the main issues, present alternative ways of addressing them under different circumstances, and caution about pitfalls to avoid. Key Words: Heuristic optimization, computational experiments 1.

The design of a cellular manufacturing system requires that a part population, at least minimally described by its use of process technology (part/machine incidence matrix), be partitioned into part families and that the associated plant equipment be partitioned into machine cells. At the highest level, the objective is to form a set of completely autonomous units such that inter-cell movement of parts is minimized. We present an integer program that is solved using a genetic algorithm (GA) to assist in the design of cellular manufacturing systems. The formulation uses a unique representation scheme for individuals (part/machine partitions) that reduces the size of the cell formation problem and increases the scale of problems that can be solved. This approach offers improved design flexibility by allowing a variety of evaluation functions to be employed and by incorporating design constraints during cell formation. The effectiveness of the GA approach is demonstrated on several problems from the literature.

Given a matrix of values in which the rows correspond to objects and the columns correspond to features of the objects, rearrangement clustering is the problem of rearranging the rows of the matrix such that the sum of the similarities between adjacent rows is maximized. Referred to by various names and reinvented several times, this clustering technique has been extensively used in many fields over the last three decades. In this paper, we point out two critical pitfalls that have been previously overlooked. The first pitfall is deleterious when rearrangement clustering is applied to objects that form natural clusters. The second concerns a similarity metric that is commonly used. We present an algorithm that overcomes these pitfalls. This algorithm is based on a variation of the Traveling Salesman Problem. It offers an extra benefit as it automatically determines cluster boundaries. Using this algorithm, we optimally solve four benchmark problems and a 2,467-gene expression data clustering problem. As expected, our new algorithm identifies better clusters than those found by previous approaches in all five cases. Overall, our results demonstrate the benefits of rectifying the pitfalls and exemplify the usefulness of this clustering technique. Our code is available at our websites.

Kohonen's self-organizing map (SOM) network is one of the most important network architectures developed during the 1980's. The main function of SOM networks is to map the input data from an n-dimensional space to a lower dimensional (usually one or twodimensional) plot while maintaining the original topological relations. A well known limitation of the Kohonen network is the “boundary effect ” of nodes on or near the edge of the network. The boundary effect is responsible for retaining the undue influence of initial random weights assigned to the nodes of the network leading to ineffective topological representations. To overcome this limitation, we introduce and evaluate a modified, “circular ” weight adjustment procedure. This procedure is applicable to a class of problems where the actual coordinates of the output map do not need to correspond to the original input topology. We tested the circular method with an example problem from the domain of group technology, typical of such class of problems. 1.

: The Vector Perturbation Approach is introduced for addressing the generalized parts grouping problem, identifying part families for a general set of suppliers, not just a single supplier. This method is driven by the need for flexible and lean supply chain systems. A vector space model is used to represent a set of operation sequences as opposed to the traditional matrix and integer programming models in Group Technology. Using this approach we find that we are able to generate part groups from 90% of the available parts, in which all the operation sequences are preserved. This contrasts with only 66% of the available parts grouped using the traditional methods. Furthermore, a vector representation of operation sequences provides an intuitive means for discovering the natural structure of the part data. From these results we conclude that this technique can dramatically improve the effectiveness of the entire supply chain. INTRODUCTION A key challenge to military readiness is the a...

The question of plant layout is central in a manufacturing process. This question becomes even more important in a mass customization context, when large product diversity has to be managed. The manufacturing process, and specifically the plant layout, has to be designed taking into account this characteristic. When all products are similar, manufacturing plant layouts are relatively easy to design; difficulties come when most products are different and require specific manufacturing operations. This paper proposes a methodology based on data mining techniques to define manufacturing plant layouts in a context of diversified products. Different steps are proposed to achieve this goal. The methodology considers: 1 / identification of representative sets of products; 2 / identification of manufacturing processes and the relevant layout (for each set of products); 3 / categorization of new products (identification of the closest set of products). The focus is on data transformations that enable to extract relevant information for the manufacturing plant layout. Key words: product families, plant layout, data transformation, data mining. 1

Cellular manufacturing emerged as a production strategy capable of solving the problems of complexity and long manufacturing lead times in batch production. The fundamental problem in cellular manufacturing is the formation of product families and machine cells. This paper presents a new approach for obtaining machine cells and product families. The approach combines a local search heuristic with a genetic algorithm. Computational experience with the algorithm on a set of group technology problems available in the literature is also presented. The approach produced solutions with a grouping efficacy that is at least as good as any results previously reported in literature and improved the grouping efficacy for 59 % of the problems. Keywords: Cellular Manufacturing; Group Technology; Genetic Algorithms; Random Keys AT&T Labs Research Technical Report TD-5FE6RN, October 29, 2002.

This paper presents a study that uses simulation to improve shop floor performance by means of two layout types and certain operational parameters. In this study, an overview of the plant layout problem is covered for the particular company. The original motivation for redesigning the entire shop floor was the need to realize improvements in material flow and output level. First, the performance of the existing system was evaluated by using ARENA. Second, manufacturing cells were formed and group technology layout was developed by means of Rank Order Clustering (ROC) method and Computerized Relative Allocation of Facilities Technique (CRAFT). Finally, the performance of the new system was evaluated and compared with that of the current system. 1

Abstract—This research was to select the optimal machine layout for laminated bamboo manufacturing by computer simulation. The laminating process was to cut bamboo trunk as laminated piece. This process was important to total total time of production line so the computer simulation was applied to run each machine layout alternatives and gather the decided parameters. Production rate (pieces/day), total time, WIP and wait time were compared for making decision. The optimal machine layout had Production rate at 12,120 12,390 laminated pieces/day and production line efficiency at 89.32%.

Inspired by principles and advantages of the group technology (GT) philosophy, part family formation for a virtual Cellular Manufacturing System (VCMS) using Fuzzy logic is designed for dynamic and uncertain conditions. In real manufacturing systems, the input parameters such as part demand and the capacity are fuzzy in nature. In such cases, the fluctuations in part demand and the availability of manufacturing facilities in each period can be regarded as fuzzy. In a dynamic environment, the planning horizon can be divided into smaller time periods where each period and/or each part has different product mix and demand. A mathematical model for virtual cellular manufacturing system as binary-integer programming is proposed to minimize the total costs consisting of fixed machine costs, variable costs of all machines and the logical group movement costs. To verify the behavior of the proposed model, a comprehensive example is solved by a branch- and-bound (B&B) method with the LINGO 12.0 software and the virtual cells(VC) are formed by defuzzification using maximizing decision level λ (lambda-cut) and the computational results are reported and compared with simulated annealing algorithm and rank order clustering algorithm.