Research On Supply Network Resilience

With the rapid growth of technology and global economy, social division becomes more elaborate and supply network comprises more companies. Today's Supply networks span the globe and involve many suppliers, suppliers'suppliers, contract manufactures, distributors, logistics providers, original equipment manufacturers (OEM), wholesalers, and retailers. This web of participating players creates more complexities, making it difficult to realize where vulnerabilities may lie. It also enhances the uncertainties of operation circumstance.Supply networks are often vulnerable to various failures. The failure which occurs at any node or link of supply network affects rapidly all companies in the web, and then, influences the normal operations of the whole network. Therefore, it has important value and significance to research on how to make supply network possess the ability of rapid recovery from failure to normal operations.We bring forward conceptual skeleton of supply network resilience quantitatively, and then set up and analyze supply network strategic design models based on resilience qualitatively. There are four parts in the paper. The central contents and main outcomes are as follows:The first part is introduction and summary of literatures.We introduce the research background, significance, contents and innovation firstly. Then, relative literatures are reviewed and appraised from four aspects and the foothold of the paper is pointed out. We put forward that supply network research has important significance to development of supply chain theories and practical operations. Current supply chain theories emphasize inter-corporation coordination and production operations in normal state and ignore the study of operations in abnormal state. More importantly, many operation ideas excessively pay attention to cost efficiency which increase supply network risk and make it hard to recover form failure. In the drastic competition times, supply chain theories focus on customer various demand and insist on matching demand deviation with supply chain flexibilities. We study the strategies of managing and recovering from disruption from the point of supply side not from demand side.The second part is about conceptual model of supply network resilience.We set up the conceptual skeleton of supply network resilience based on many correlative cases. Three questions are answered: what is supply network resilience? Why study supply network resilience? And how to increase supply network resilience? The part provides theory foundation for latter mathematical models.Supply network resilience is defined as the ability of supply network to recover to normal state from disruption impact. Here, disruption is confined within high impact/low probability events. Resilience is the ability to recover and is not the ability of anti-impact and adapting impact. This is the difference for resilience with robustness and flexibility. We also suggest that supply network vulnerability can be studied from two dimensions: likelihood and consequences. Then, disruption events are researched form the two points. We analyze the factors which induce supply network vulnerabilities.We consider that designing build-in resilience is the most important way to improve supply network resilience. Four principles are brought forward. First, efficiency and redundancy should be traded-off which involves setting up strategic emergency stock, designing redundant capacities and adopting options supply. Second, multi-sourcing supply strategy is used. Third, resilience should be considered in designing distribution network and channel. Finally, information system design should include resilience. For a single manufacture company, resilience can be improved from production flexibility. We also advance the idea of transferring from JIT to JIC. Supply network relations management and building a culture of risk are also the important means to improving resilience.The third part is about mathematical models on supply network design. The part is divided into two segments: inbound supply network design and outbound supply network design.Inbound supply network design includes multi-sourcing model and contingency supply model. Supplier sourcing limit parameter is mounted in multi-sourcing model to achieving multi sourcing. The model also considers supplier reliability and expected tolerable service level. The results show that multi-sourcing can improve supply network resilience; moreover, trading off between resilience and cost can be got by adjusting expected tolerable service level. Contingency supply model bases on multi-sourcing model and involves supply network reliability. Strategic emergency stock and options supply are introduced as two contingency supply modes in the model. The simulations results reveal that contingency supply model can significantly improve network resilience and reduce disruption lose, but it increases operation cost. Therefore, we should design appropriate model parameters to get sustainable cost and best resilience.Resilient outbound supply network model are established based on PMM and UFLM. In the model, we consider not only single distribution center failure but also multi distribution center simultaneous failures. The principle of modeling is that each customer is assigned to a primary facility that will serve it under normal circumstances and a backup facility (or a set of backup facilities) that will serve it when the primary facility has failed. Maximal tolerable failure cost is set up as resilience constraint. The value of the model reveals by decision curve. Decision curve can tell us that we can find optimized outbound supply network design once the maximal tolerable failure cost is determined. Hence, managers only need to decide the value of maximal tolerable failure cost according to their personal risk appetites and the aims of designing the network.The last part is the conclusion and future prospect.The part summarizes the main views and indicates the defects as well as advices and expectations of future work.The innovation in the dissertation includes four aspects. These are as follows:1. The research on supply network failure is absent in current literature, whereas the dissertation just base on it. The paper brings forward considering resilience in designing supply network which challenges the traditional supply network design principle which focuses on cost efficiency. Other than most papers studying supply chain from demand side, the paper researches on supply disruption from supply side. This is the new research view.2. Conceptual research skeleton on supply network resilience is brought forward which establishes base for resilience study. Supply network resilience is restricted as the ability of supply network to recover to normal state from the consequences of high impact/low probability disruptions. The differences of resilience with robustness and flexibility are analyzed which amends the confusion in some literature. We also suggest that supply network vulnerability can be studied from two dimensions: likelihood and consequences. Five design principles and operation ideas are brought forward to improve supply network resilience.3. Resilient inbound supply network models are set up which make up the lack of quantitative research on resilience. We build multi-sourcing model and then models contingency supply which includes strategic emergency stock and options supply. The three design principles can commendably improve supply network resilience. The paper also points out that it is not unlimited to improve resilience. The tradeoff between cost and resilience must be considered in supply network design.4. Resilient outbound supply network models are set up. Most former resilience researches aim at supplier failure and the outbound supply network resilience is absent. So this is the good makeup. The principle of modeling in the paper is that each customer is assigned to a primary facility that will serve it under normal circumstances and a backup facility (or a set of backup facilities) that will serve it when the primary facility has failed. This has essential difference with traditional design principle which only assigns a facility for each customer and only considers cost factor. The decision curve can makes designer to achieve a ideal supply network by adjusting model parameters according to real instance.