Comparison of a visual analytics simulation with real world manufacturing lines

Abstract

Global competition between manufacturing companies forces them to offer ever more customizable products to their clients, while keeping high production levels. To assure companies thrive in this environment, they store large amounts of data from their production facilities and conduct simulations in order to optimize their operations. Analysts need to use techniques to deal with large amount of data efficiently and without being overwhelmed by it. Visual Analytics handles exactly that challenge. The goal of this thesis is to develop an approach that help the user gain knowledge that is useful for conducting simulation studies faster. In this thesis, a visualization assists analysts in calibrating and validating settings using historical data. The Critical Path Method was successfully applied to precisely trace bottlenecks in a simulated manufacturing line. In addition, new metrics were developed to guide the optimization of schedules where the analyst may control the performance of resources over an entire sequence of jobs, but not any of them individually, such as in manufacturing machines. The important metrics are visualized and the sequence of operations was displayed by an interactive lens.