Operating an effective resource allocation for wire bonders in the semiconductor assembly industry

Taiwan is the global leader in capacity and technology in the semiconductor testing and assembly industry; its firms must confront the tremendous competitive pressure from more than 30 other related firms in Taiwan and a market force for vertical integration of the semiconductor manufacturing industry. One of the key issues to a firm's success is how to utilise its semiconductor assembly and testing resources against dynamic demands. This study proposes an effective resource allocation approach for wire bonders from the perspectives of comparative advantage and opportunity cost, and implements it using a decision tree to match the complex product requirements with different levels of bonder resource availability. There are two objectives: (i) when the capability of resources exceeds the demand of products, the proposed approach aims at keeping higher-level resources as unencumbered as possible; and (ii) when resources are insufficient to meet the demands, the approach aims at minimising computation time for a resource allocation plan, making full utilisation of resources. Experimental results from a leading semiconductor assembly firm have shown that the proposed model outperforms its conventional approach in terms of resource capability appropriation and computational time.