Abstract: The semiconductor wafer fabrication system is one of the most complex discrete manufacturing systems owing to its great number of production steps, heavy re-entry production flow, various kinds of products, etc. The automatic material handling system plays a key role in improving the production efficiency, reducing the work in process (WIP), and shortening the production cycle time of the semiconductor factory. To rapidly and effectively evaluate the impact of crossover systems on the overall performance of transport systems in wafer production, a performance analysis model was built by mathematical analysis and a decision-making method of crossovers was established. In the modeling procedure, the retrial transportation rule was first introduced. Then, considering the change in retrial rate affected by the crossovers strategy and referencing the related retrial queuing model, the mathematical expression of the expected WIP was constructed. Simultaneously, according to the cost of each crossover, the mathematical expression of the cost of crossovers was built. The optimization analysis model was obtained with the objective of minimizing the number of crossovers and WIP. Furthermore, the constraint conditions based on the length of the conveyor belt were introduced, so that the number of WIP products on each conveyor belt and crossover would not exceed the capacity limit of the conveyor belt. Finally, the NSGA-Ⅱ algorithm was used to solve the multi-objective optimization problem. The relation between the number of WIP and cost of crossovers under different logistics loads was investigated, and Pareto frontier charts were drawn up for comparative analysis. The influence of parameter settings on the objective function of the system was studied, and the solution set was classified by clustering the analysis algorithm. Based on the aboveanalysis results, the best strategy of system crossovers was summarized. By comparing the proposed system with the existing system and transportation strategy, the effectiveness of the strategy in reducing cost and WIP was confirmed.