In this study three different methodologies are developed, (i) a similarity coefficient based approach is adopted to obtain the initial machine part grouping solution. Thereafter a state-of-the-art part clustering algorithm is proposed to improve the solution quality. (ii) A dissimilarity coefficient based approach is adopted to obtain the initial machine part grouping solution. Thereafter a novel approach based on a population based heuristic technique is proposed to improve the solution quality. (iii) A similarity coefficient based technique is exploited to form machine cells and part families. Thereafter a Simulated Annealing (SA) based metaheuristic algorithm is utilized to enhance the quality of the solutions obtained.