A genetic algorithm tool for optimising cellular or functional layouts in the capital goods industry

The literature on the design of manufacturing facilities has two major themes: the application of cellular manufacturing, including the use of clustering methods; and the solution of facilities layout problems using optimisation methods. Most previous research has been based upon relatively small or theoretical problems. This paper presents a Genetic Algorithm optimisation method that has been developed which can be applied to a set of manufacturing cells or to an entire manufacturing facility. The approach can be used for either green field or brown field layout problems. The model was tested using a large data set from a collaborating capital goods company. Genetic Algorithm programs include a number of parameters including the probabilities of crossover and mutation, the population size and the number of generations. A full factorial experiment was performed to identify the best configuration. The results were compared with the Company's layout and the best layout that could be generated randomly. When the layout was considered as brown-field problem there was a reduction of total rectilinear distance travelled of 25% compared to the Company's layout. The number of generations was the only statistically significant factor. When the layout was treated as a green-field problem the total rectilinear distance travelled was reduced by 70% and the population size, the number of generations and the probability of crossover were statistically significant.