Högskolan i Skövde

Many simulation-based optimization packages provide powerful algorithms to solve large-scale system problems. But most of them fall short to offer their users the techniques to effectively handle decision variables that are of multiple-choice type, as well as equality constraints, which can be found in many real-world industrial system design and improvement problems. Hence, this paper introduces how multiple choice sets and Manhattan-distance-based constraint handling can be effectively embedded into a meta-heuristic algorithm for simulation-based optimization. How these two techniques have been applied together to make the improvement of a complex production system, provided by an automotive manufacturer, possible will also be presented.

Bottleneck is a common term used to describe the process/operation/person that constrains the performance of the whole system. Since Goldratt introduced his theory of constraint, not many will argue about the importance of identifying and then improving the bottleneck, in order to improve the performance of the entire system. Nevertheless, there exist various definitions of bottleneck, which make bottleneck identification and improvement not a straightforward task in practice. The theory introduced by Production Systems Engineering (PSE) that the bottleneck of a production line is where the infinitesimal improvement can lead to the largest improvement of the average throughput, has provided an inspirational and rigorous way to understand the nature of bottleneck. This is because it conceptually puts bottleneck identification and improvement into a single task. Nevertheless, it is said that a procedure to evaluate how the efficiency increase of each machine would affect the total performance of a line is hardly possible in most practical situations. But is this true?In this paper, we argue how multi-objective optimization fits nicely into the theory introduced by PSE and hence how it can be developed into a practical bottleneck improvement methodology. Numerical results from a real-world application study on a highly complex machining line are provided to justify the practical applicability of this new methodology.

Manufacturing companies of today are under pressure to run their production most efficiently in order to sustain their competitiveness. Manufacturing systems usually have bottlenecks that impede their performance, and finding the causes of these constraints, or even identifying their locations, is not a straightforward task. SCORE (Simulation-based COnstraint REmoval) is a promising method for detecting and ranking bottlenecks of production systems, that utilizes simulation-based multi-objective optimization (SMO). However, formulating a real-world, large-scale industrial bottleneck analysis problem into a SMO problem using the SCORE-method manually include tedious and error-prone tasks that may prohibit manufacturing companies to benefit from it. This paper presents how the greater part of the manual tasks can be automated by introducing a new, generic way of defining improvements of production systems and illustrates how the simplified application of SCORE can assist manufacturing companies in identifying their production constraints.

Manufacturing companies are operating in a severely competitive global market, which renders an urgent need for them to explore new methods to enhance the performance of their production systems in order to retain their competitiveness. Regarding the performance of a production system, it is not sufficient simply to detect which operations to improve, but it is imperative to pinpoint the right actions in the right order to avoid sub-optimizations and wastes in time and expense. Therefore, a more accurate and efficient method for supporting system improvement decisions is greatly needed in manufacturing systems management. Based on research in combining simulation-based multi-objective optimization and post-optimality analysis methods for production systems design and analysis, a novel method for the automatic identification of bottlenecks and improvement actions, so-called Simulation-based Constraint Identification (SCI), is proposed in this paper. The essence of the SCI method is the application of simulation-based multi-objective optimization with the conflicting objectives to maximize the throughput and minimize the number of required improvement actions simultaneously. By using post-optimality analysis to process the generated optimization dataset, the exact improvement actions needed to attain a certain level of performance of the production line are automatically put into a rank order. In other words, when compared to other existing approaches in bottleneck detection, the key novelty of combining multi-objective optimization and post-optimality analysis is to make SCI capable of accurately identifying a rank order for the required levels of improvement for a large number of system parameters which impede the performance of the entire system, in a single optimization run. At the same time, since SCI is basically built a top a simulation-based optimization approach, it is capable of handling large-scale, real-world system models with complicated process characteristics. Apart from introducing such a method, this paper provides some detailed validation results from applying SCI both in hypothetical examples that can easily be replicated as well as a complex, real-world industrial improvement project. The promising results compared to other existing bottleneck detection methods have demonstrated that SCI can provide valuable higher-level information to support confident decision-making in production systems improvement.

Many simulation-based optimization packages provide powerful algorithms to solve industrialproblems. But most of them fail to oer their users the techniques they needto eectively handle multiple-choice problems involving a large set of decision variableswith mixed types (continuous, discrete and combinatorial) and problems that are highlyconstrained (e.g., with many equality constraints). Yet such issues are found in manyreal-world production system design and improvement problems. Thus, this paper introducesa method to eectively embed multiple choice sets and Manhattan-distancebasedconstraint handling into multi-objective optimization algorithms like NSGA-II andNSGA-III. This paper illustrates and evaluates how these two techniques have been appliedtogether to solve optimal workload, buer and workforce allocation problems. Anexample follows, showing their application to a complex production system improvementproblem at an automotive manufacturer.

By innovatively formulating a bottleneck identication problem into a bi-objective optimization,simulation-based multi-objective optimization (SMO) can be eectively used as a new method for gen-eral production systems improvement. In a single optimization run, all attainable, maximum throughputlevels of the system can be sought through various optimal combinations of improvement changes ofthe resources. Additionally, the post-optimality frequency analysis on the Pareto-optimal solutions cangenerate a rank order of the attributes of the resources required to achieve the target throughput levels.Observing that existing research mainly put emphasis on measuring the convergence of the optimizationin the objective space, leaving no information on when the solutions in the decision space have convergedand stabilized, this paper represents the rst eort in increasing the knowledge about the convergence ofSMO for the rank ordering in the context of bottleneck analysis. By customizing the Spearman's footruledistance and Kendall's tau, this paper presents how these metrics can be used eectively to provide thedesired visual aid in determining the convergence of bottleneck ranking, hence can assist the user todetermine correctly the terminating condition of the optimization process. It illustrates and evaluatesthe convergence of the SMO for bottleneck analysis on a set of scalable benchmark models as well as twoindustrial simulation models. The results have shed promising direction of applying these new metrics tocomplex, real-world applications.

Bottleneck analysis can be defined as the process that includes both bottleneck identification and improvement. In the literature most of the proposed bottleneck-related methods address mainly bottleneck detection. By innovatively formulating a bottleneck analysis into a bi-objective optimization method, recent research has shown that all attainable, maximized TH of a production system, through various combinations of improvement changes of the resources, can be sought in a single optimization run. Nevertheless, when applied to simulation-based evaluation, such a bi-objective optimization is computationally expensive especially when the simulation model is complex and/or with a large amount of decision variables representing the improvement actions. The aim of this paper is therefore to introduce a novel variables screening enabled bi-objective optimization that is customized for bottleneck analysis of production systems. By using the Sequential Bifurcation screening technique which is particularly suitable for large-scale simulation models, fewer simulation runs are required to find the most influenacing factors in a simulation model. With the knowledge of these input variables, the bi-objective optimization used in the bottleneck analysis can customize the genetic operators on these variables individually according to their rank of main effects with the target to speed up the entire optimization process. The screening-enabled algorithm is then applied to a set of experiments designed to evaluate how well it performs when the number of variables increases is a scalable, benchmark model, as well as two real-world industrial-scale simulation models found in the automotive industry. The results have illustrated the promising direction of incorporating the knowledge of influencing variables and variable-wise genetic operators into a multi-objective optimization algorithm for bottleneck analysis.