Selecting material feeding system of the assembly line by means of artificial neural network method

Manufacturing companies which aims to become preffered in competitive markets should decrease cost and lead time while increasing quality and efficiency. Material feeding system is one of the important parameter which effects the efficiency. Manufacturing companies should take into consideration of some criteria such as, great variety of materials, material volume, weights, product demands, area requirements in facility, labor force etc. to decide the most suitable material feeding method. In this study, material feeding method decision making problem is discussed in a manufacturing company, produces hydraulic pump for material feeding to an assembly lines. The aim of the study is to determine the most efficient material feeding method of existing materials for assembly lines in different value streams in the plant and to easily decide on the feeding method of new materials. A feedforward neural network structure is used to solve material feeding method selection problem in this study. Input and output criteria are determined according to literature research and stratgeic decisions of company. Multilayer artificial neural network topolgy and back propogation algorithm is applied for complex problem. Kitting method is proposed which is not currently applied. The proposed approach provides an effective solution for decision making to choose the most suitable material feeding system and this method can easily be used for the new materials in new assembly lines. According to the implementation results of the proposed approach increase in labor productivitiy and quality, improvement of ergonomic conditions, decrease for empty space requirement, necessary time for planning and decision making was observed. .