Global agriculture trade stimulates a transformation of traditional cold chains to global cold chains with long-haul intercontinental transportation. Sea transportation by reefer containers will be the dominant modality in the future for global cold chains. Reefer logistics plans and controls the forward and backward logistics of reefer containers and perishable products. Reefer logistics is challenging due to its distinctive features that include (1) high asset value, (2) food quality/waste issue, (3) energy requirements for refrigeration, (4) extra emissions, and (5) reefer maintenance and PTI. Thus, reefer logistics is a complex system with specific technical challenges, which deals with cost-efficiency, timeliness, product quality, and sustainability requirements.
One of the biggest challenges in reefer logistics is to find solutions balancing cost, quality, and sustainability since improving product quality and sustainability may come at a cost. Additionally, multiple stakeholders are involved in reefer logistics, who have different perceptions of the product quality, economic aspect, and sustainability. There can be a conflict of interests between stakeholders. Therefore, it is necessary to manage the potentially conflicting objectives in a multi-actor setting for reefer logistics. The main research question that this thesis aims to answer is: How to manage reefer logistics considering its distinctive characteristics and trading off conflicting objectives of multiple stakeholders?
In this thesis, we firstly provide an overall understanding of reefer logistics by analysing its technical characteristics and stakeholders. After the qualitative analysis, we propose a generic agent-oriented modelling framework to manage reefer logistics, which considers the identified features and the multiple stakeholders' objectives. We further evaluate two (re)designs of reefer logistics using the overall understanding of reefer logistics and the generic modelling framework.
In Chapter 2, we explore the characteristics of reefer logistics. We carry out a systematic literature review on reefer logistics and conduct system analysis based on the literature review and expert interviews. We first focus on single-actor system analysis. We identify the means, objectives of each actor, and the internal/external factors affecting the achievement of objectives. Next, we conduct a multi-actor system analysis to study how an action (carried out by one actor) could influence other chain actors. The results of system analysis in chapter 2 can be use as an explanatory tool for further development in decision support models for reefer logistics.
After analysing the reefer logistics system, we further looked into the decision-support modelling for reefer logistics considering its specific characteristics and different objectives of stakeholders. In Chapter 3, we develop an agent-based simulation (ABS) framework for reefer logistics. The ABS trades off operational cost, quality, and CO2 emission, which can support decision-making in the design and operation of (global) reefer logistics systems. The ABS is developed based on the knowledge acquired in Chapter 2, which specifies the main actors, resources, activities, and flows. We model a reefer logistics system with a two-tier architecture that consists of a social layer and a physical layer. On the social level, each actor as an autonomous agent makes decisions and interacts with other actors. On the technical level, physical components perform activities and handle reefer/cargo flows. Next, we carry out a numerical case study of a global banana supply chain. The case study shows the capability of the ABS model to compare scenarios with different cold chain configuration designs and management strategies in terms of operational cost, CO2 emissions, product quality, and waste.
In Chapter 4, we explore the capability of the ABS model developed in Chapter 3. The ABS model is applied to evaluate several possible (re)design scenarios enabled by the remote container management system (RCM). We first developed a conceptual framework to assess the value of RCM. In the framework, we classify the RCM information into four categories – (1) product and transportation condition, (2) security, (3) container location, and (4) container status. The value of RCM information is from two aspects: (1) substitution of existing information flow as a new technology, and (2) decision-support with new types of information. Secondly, we did a numerical case study of a global banana supply chain to quantify the value of RCM. We evaluate five information scenarios that focus on both IT and decision support aspects. We can conclude that RCM enables digitalization that has an improvement in the efficiency of reefer logistics. The beneficiaries are mainly the cargo owner (shippers) and the asset owner (shipping lines and leasing companies). Other actors may also benefit from the RCM information, such as customs and terminal operators.
In Chapter 5, we looked into the concept of flow consolidation in port-hinterland container transport. We develop an analytical model to compare the difference in implementing flow consolidation scenarios between reefer logistics and dry container logistics. The model compared the performances in various dimensions of three flow consolidation scenarios of hinterland transportation for both dry and perishable cargoes: (1) only-trucking, (2) container consolidation, and (3) combined container/cargo consolidation. We derived several propositions and theorems describing conditions under which different scenarios outperform others regarding operational cost and CO2 emission. We further conducted a numerical case study that shows cargo type and shipment distance are essential factors that affect flow consolidation performances.
In this thesis, we develop explanatory and modelling tools for reefer logistics to trade off different objectives, e.g., cost, product quality, and emission. We summarize the findings in five main concluding statements: (1) Trading off different objectives is essential when making decisions for reefer logistics. (2) The goals of multiple actors in cold chains can be conflicting. Therefore, integrating multiple actors’ viewpoints is essential in finding (and consequently accepting) global optimal solutions for reefer logistics considering their (conflicting) goals and market positions. (3) A generic agent-based simulation model is an appropriate way to investigate different what-if scenarios to find tailored solutions for reefer logistics. (4) Traceability and transparency are crucial for reefer logistics to improve the safety/security of perishable products and the cost-efficiency and sustainability of logistics processes. (5) Traditional solutions for supply chain management might not be necessarily suitable for cold chains due to the technical complexity of cold chains.