1.2 Synchromodal transport planning with demand uncertainties (DISPATCH project)

Session 1.2 |  13th of June from 13.30 to 15.00 (Athens local time UTC + 3) 

Organizer: Dispatch Project 

Agenda and presentations: 

• Synchromodal inventory replenishment under non-stationary demand with a partially observable demand process. Hannah Yee. PhD student. KU Leuven 

In this paper, we study synchromodal inventory replenishment under non-stationary demand using road and rail transport. Synchromodal inventory replenishment means that transport decisions consider real-time information about the need for inventory replenishment that is driving the transport orders. Non-stationarity is modeled through changes in the underlying demand distribution, yet only partial information about it is available to the decision-maker. We propose a policy combining a pre-committed base order on rail transport with flexible short-term orders on both road and rail transport. The flexible short-term orders consider the inventory needs and partial information about the non-stationary demand. We demonstrate the policy in a numerical experiment with replenishment shipments from a supplier in Spain to an inventory center in Belgium. In the experiment, we analyze the impact of having flexible rail orders in our policy. The results reveal that the use of flexible rail orders reduces both costs and emissions for a shipper and increases revenue for a rail operator. 

• Synchromodal transport: How are the benefits of collaboration distributed? Javier Durán-Micco. Postdoc researcher. Vrije Universiteit Brussels 

Synchromodality is a novel concept that aims for more sustainable logistic operations in dynamic environments. As PI, it refers to multimodal logistic networks in which decisions about routing and mode shift are taken in real-time. This implies a strong collaboration between actors, who would be required to share information and capabilities. Therefore, before a potential implementation, it is necessary to show the impacts of synchromodal operations on individual players. This study uses a simulation model to analyze the impact of different scenarios in the operations and revenues of different actors within a multimodal logistic network. We focus on the interaction between logistic service providers (LSPs) or between LSPs and carriers. We assume that LSPs have certain capabilities, such as truck fleets and booked capacity in train services, which can be used to handle transportation requests. Moreover, they can provide or requests additional resources to other LSPs or directly to the carrier, to satisfy unexpected demand or react to disruptions. Different scenarios are tested, considering different collaboration schemes or variable relative size between the players. The results show how costs and revenues are distributed in each case, and provide insights on what incentives are needed to increase fairness in the system. 

• Long-term capacity planning in rail-road networks under demand uncertainty. Thibault Delbart. PhD Student. Hasselt University, Logistics Research Group 

Intermodal transport combines road transport with high-capacity transport modes, which is potentially more efficient than unimodal road transport. Logistics service providers (LSPs) typically buy transport capacity from third parties, which is often done long in advance on high-capacity transport modes. This leads to uncertainty regarding both the demand and the remaining available capacity at a later time. In literature, a common approach to assist such capacity decisions is with two-stage models that wait until complete demand information is available before updating capacity. In reality, LSPs update their transport capacity gradually as new information becomes available. This research proposes a more realistic decision support model to assist capacity planning in a rail-road network from the perspective of an LSP. It contains an additional stage during which transport capacity is updated with partial information. Included types of uncertainty are the demand volume and the amount of transport capacity that is available on the market. The performance of our model is compared against a two-stage model over a set of theoretical instances with a factorial design. 

•    Synchromodal transport re-planning using Agent-Based Modeling. Shafagh Alaei. PhD researcher. Vrije Universiteit, Brussels

In recent years, there has been a growing interest in Synchromodal Transport (ST), which aims to optimize the use of different modalities to reduce costs, improve reliability, and enhance sustainability. The traditional approach of transport planning, in which each modality is operated independently, often leads to inefficiencies. ST aims to address this issue by integrating different modes, allowing for the optimization of each mode and the overall supply chain. However, ST is often subject to disruptions and changes in demand, which require the re-planning of the itineraries. This paper presents an agent-based simulation model for ST planning, with a focus on re-planning under disruptions. Using Agent-Based modeling, we study the behavior of multiple actors involved in an ST system and their interactions, as well as the impact of their behavior on the network. The model explores the ST network at the operational level and from the Logistics Service Providers’ (LSPs') perspective. A numerical experiment is conducted to evaluate cost savings and emissions reduction under different collaboration and re-routing scenarios. Our findings show that synchromodal scenarios lead to higher flexibility and reliability than business-as-usual scenarios. Moreover, our model verifies that the cost saving is considerable when LSPs collaborate with each other.