Abstract
Abstract
Today, many organizations are following traditional manufacturing layouts. These are easy to form, but involve very high overall manufacturing cost. An alternative is to introduce a cellular manufacturing system in modern production shops, which recognizes the existence of machining groups and their dedicated component families in discrete lot manufacturing. The essential problem in implementing cellular manufacturing is in the identification and formation of the machine cells. This paper proposes a new methodology for cell formation utilizing a syntactic pattern recognition approach. The selection of an appropriate cell for a new part is based on the operational information of the part. With the use of self-organizing neural networks, the machine cells are identified. Results are presented with a numerical example.