Abstract
The current paper proposes an optimum clustering approach for automatic generation of machine cells and part families. The design of a cellular manufacturing system begins with a specified 'part-machine incidence' matrix, showing the machine sequence and volume of production. The arrangement of machines in the cells is addressed in the present paper by finding an optimal machine sequence, which maximizes the overall flow of components between the machines. An optimal number of cells is arrived heuristically by accounting the total intercell movements using the obtained machine sequence. The methodology is illustrated with examples.