A stochastic logistics model for Indonesia's national freight transport model: Transport chain choice from the shipper perspective

This paper presents research towards the development of a stochastic approach for estimating the transport chain choice for domestic shipments in Indonesia. This stochastic model aims to improve the logistics choices within Indonesia's national freight transport model (INTRAMOD), which currently handles such choice deterministically. The INTRAMOD logistics model presents five distinct transport chain possibilities involving four main modes: truck, rail, ship, and plane. To acquire the necessary data, revealed preference (RP) and stated preference (SP) survey work has been undertaken. Using the obtained RP/SP data, multinomial logit (MNL) models have been used to estimate the transport choice model. The model with a single time coefficient was found to be superior to the other models. Additionally, this preferred MNL model was extended by segmenting according to shipment characteristics, particularly for high and low value of goods. The results indicate that shipments with a high value of goods are more sensitive to transport time.