Shipment planning with uncertainties in global logistics management

Abstract

Global logistics management has been playing a critical role in modern international business. Airfreight forwarders are third-party logistics service providers responsible for handling the majority of tonnage in international cargo shipping processes. Good shipment planning, which involves optimizing orders, coordinating shipment activities and arranging available transportation resources, is required to achieve both shipment reliability and cost efficiency. In practice, cargo shipments are exposed to various unforeseen risks. Factors such as bad weather, natural disasters, customs clearance tardiness, carrier overbooking, and system malfunction may cause uncertainties in ordering quantity and shipment reliability. Thus, shipment planning should consider the different types of uncertainties possibly existing in international cargo shipping processes. In this thesis, we focus on the study of shipment planning with uncertainties in cargo shipping processes. Two research problems are presented. First, we consider a shipment planning problem with random activity processing times at the early shipping stage. We propose a two-stage stochastic shipment planning approach to the production of a proper shipment plan considering recourse treatment. Second, we consider a replenishment problem for real-time monitoring a cargo shipping process from the retailer's perspective. Real-time information on market demand and shipment status is used to determine the optimal ordering strategy for the retailer. The first part of our study addresses the shipment planning problem with uncertainties for an airfreight forwarder responsible for multiple shipments in the planning period. Shipment activities are divided into two stages based on regional settings or processing sequence. Activity processing times in stage 1 are assumed random, whereas those in stage 2 are considered deterministic. At the beginning of cargo shipping processes, the forwarder determines a start-up shipment plan and acquires the resources used in shipping processes. In case of individual shipment tardiness occurring in stage 1, the airfreight forwarder may implement an in-process adjustment on the existing shipping arrangement to improve shipment reliability and reduce costs. Since shipment reliability - the probability of delivering a shipment on time - is regarded as one of the most important performance measurements in shipping process, a good shipment planning policy should consider both shipment cost and reliability efficiency. To emulate the stochastic problem for profit maximization, we establish a two-stage stochastic shipment planning model using a scenario generation approach. A two-level optimization mechanism is designed for obtaining the optimal start-up shipment plan. We present an illustrative example with industry data in South China. In the part of the managerial analysis, performances of the optimal shipment plans obtained from the two-stage stochastic shipment planning approach and the deterministic model are examined. The specific effects of target delivery times, variability of activity processing times, adjustment costs, and penalty costs are investigated. Our proposed two-stage stochastic shipment planning model tends to handle the random activity processing times better than the deterministic shipment planning model does. The second part of our study considers a real-time replenishment problem with random demand and shipping conditions for a retailer. Apart from demand fluctuation, shipping conditions of an outstanding order involving the shipping quantity and shipping time should be considered in the design of replenishment policies. At the beginning of a regular ordering cycle, the retailer takes a regular order from the supplier. During the cargo shipping process, the retailer may trigger an optional order to improve the expected profit by comparing the profit value in the adjustment case with the one in the no-adjustment case. Adopting the real-time statistical process control (SPC) concept, we derive the optimal replenishment policies in different conditions of the shipping process. Comparative analyses are conducted for replenishment policies determined under different levels of information disclosure. The effects of different types of information on decisions and managerial insights are discussed. With the use of more updated information during the shipping process, our proposed replenishment policy indicates improved performance over traditional models when only the inventory uncertainty is considered.

Date of Award	15 Jul 2013
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Yer Van HUI (Supervisor)