An Agent-based Modeling Approach to Agile Supply Network


    The Dongdaemun fashion apparel market in Seoul has formed a unique supply network through evolutionary process. The market constituencies include thousands of retailers and wholesalers, intermediaries and raw materials suppliers, and manufacturers. Without any central control, the market has managed to become one of the most efficient and flexible suppliers of apparels in the world. The latest fashion trends from Paris, New York, and Tokyo are reflected in the display shelves within days. Replenishment lead time of hot-selling items is only one or two days. Average product life cycle of an item is two weeks from design to salvaging. In addition to this lightening speed, the quality of products is good and prices are relatively cheap. Looking inside the market, we found that the roles and behaviors of the players have changed continuously like any other complex adaptive systems.
    In this talk, we focus on the two important dynamics of the market. One is how the market adapts to the changing fashion trends by imitation, learning, and innovation. The other is how the risk sharing (via return policy) is determined between wholesalers and retailers when facing the demand uncertainty. We report the early results from simple agent-based simulation models. Initial insights from the experiments and future modeling directions will be discussed.


Seung-Kyu Rhee
KAIST Graduate School of Management
http://kgsmweb.kaist.ac.kr
skrhee@kgsm.kaist.ac.kr