An approach to machine group and part family formation in reconfigurable manufacturing systems

In response to the rapidly changing market demand, manufacturing firms need to adopt continuous and rapid changes in the manufacturing strategies to gain a competitive edge. In addition, there is an increasing trend for achieving higher level of integration between the design and manufacturing function to make manufacturing more efficient and productive. Reconfigurable Manufacturing System (RMS) stands as an effective solution towards achieving this goal. RMS is a manufacturing system which achieves the required configuration in its structure for rapidly adjusting the production capacity and functionality within its part family in response to sudden changes in market demands. The formation of part families is one of the core issues for designing the RMS. To achieve this objective, the concepts of Cellular Manufacturing are utilized for the grouping machines into manufacturing cells and parts into their families. As for the representation of the cell formation problem is concerned, most of the researchers use '0' -- '1' binary machine part incidence matrix. This research makes use as many as real life production information in the input matrix for representing the cell formation problem. The objectives of minimizing the cell load variation and intercellular flows are considered for the cell formation problem. The methodology uses Genetic Algorithms to group machine into the manufacturing cells and products into their families. The machine utilization capacity is calculated for random demand changes. The algorithm is coded in Java language and is run on the Eclipse platform.