On genetic algorithms for shoe making nesting - A Taiwan case

This paper proposes a methodology that integrates in-house placement heuristics with genetic algorithms to solve the nesting problems of shoe making. The problems are classified as placing a set of irregular patterns on a regular area and limited to at most two different types of patterns on the area. Because of the intractability of the nesting problem, our objective is to utilize genetic algorithms' fast convergence and solution quality to improve material utilization and reduce the calculation time of the pattern. Using the real-life data of two international brands of athletic shoes, the empirical results show that our proposed methodology can reduce average material requirements by 2.64% and average nesting time by 69.15% compared to those of current in-house software. The reduction of materials is becoming more important given that the industry is facing continuingly declining profit margins.