Search

At Kunming International Flower Auction Market (KIFA), about 2.5 million cut flowers traded in 10,000 transactions need to be distributed daily to buyers in its distribution area. Small lots and many buyers per trolley are two distinctive features at KIFA and the identities of the buyers and their demands are not known in advance. The growing transaction volume has recently increased the distribution workforce and the buyers’ waiting time. In this paper, we introduce a modified class-based location policy using KIFA’s historical data to improve its current put system performance. We use the closest-open location method in each class area, which improves the put system performance at KIFA. We examine the effects of the distribution area shape and the number of blocks in each class area on performance measures, and find that KIFA’s put system performance can be further improved. [ABSTRACT FROM AUTHOR]