ATO supply chains within the context of system dynamics

2.3.1. System dynamics studies in the context of the CODP, MTO and ATO systems

An important part of the ATO system is the position of CODP as well as the MTO phase, in which the latter is characterised by the order-driven, customized-centric operations environment. This is completely different from the MTS environment where tangible inventory plays the key role in influencing supply chain dynamics. Given that extensive MTS dynamic studies has been presented in literature (reviewed in Section 2.2), this section exclusively focuses on literature considering MTO, CODP and, the focus of this thesis,. the ATO system

Wikner et al. (2007) developed an MTO system dynamics model and explore its dynamic performance by using the order book feedback control concept. They suggested that managers may be able to control the level of capacity and lead time flexibility by selecting appropriate forecast smoothing and order book control parameters. The limitation of this work, however, is the ignoring of nonlinearities presented in the MTO system, and also, although the model could potentially be extended and used for the dynamic analysis of decoupled systems, it lacks a mechanism for integration between the MTS and MTO elements

Özbayrak, et al. (2007) developed four-echelon MTO based system dynamics model and analysed some key dynamic metrics such as inventory, WIP levels, backlogged orders and customer satisfaction. Although some insightful dynamic results were obtained and analysed, the pure simulation approach lacks the analytical power in giving guidance regarding the improvement and engineering of the supply chain system.

Literature review

45

Anderson et al. (2005) assessed the dynamic performance of order-based service supply chains with different degrees of demand variability and information sharing. They developed a capacity management model for a serial chain by presenting related capacity, processing, backlog and service delays at each supply chain stage. By using the system dynamics simulation approach, they characterise the bullwhip phenomenon exist in such supply chain systems. The impact of different levels of information sharing and management strategies on capacity and service delay variability are also studied

By decoupling generic FD and CD models, Hedenstierna and Ng (2011) evaluated the dynamic consequences of shifting the position of the CODP and found that the ideal position depends on the frequency of demand. However, their model is simple and linear, lacking more realistic representations, such as capacity constraints and availability of material. Choi et al.’s (2012) developed a system dynamics simulation model from Lee and Tang’s (1997) model and their experiences gained through a case study in a Korean automobile manufacturer. In contrast to Hedenstierna and Ng (2011), their model represents complex variable relationships, but their simulation results are limited to Korean global automobile companies.

Wikner et al. (2017) conceptually develop a hybrid MTS-MTO model that can represent a typical ATO system by decoupling the customer orders at the final assembly plant. By using system dynamics simulation, they highlight the significant impact of capacity constraint downstream of the CODP on backlog and CODP inventory dynamics, although the conceptual model does not explicitly consider the upstream capacity limit as well as the delivery LT measurement. Since this study focuses on the ATO system within the context of

Literature review

46

the semiconductor and PC industries, the corresponding relevant background and literature related to the system dynamics are now introduced.

In general, few studies have investigated the MTO, CODP and ATO system structure from the system dynamics perspective. The nature of most studies is conceptual without support of real-world context. Also, most studies adopted a pure system dynamics simulation approach and this leads to the difficulties in obtaining analytical insights regarding the control and design of appropriate policies in improving supply chain dynamics performance.

2.3.1. ATO system dynamics in the PC industry

Very limited effort has been found for modelling and analysing the system dynamics of the ATO system structure in the PC sector. Berry and Towill (1992) developed causal loop diagrams to explain the ‘gaming’ that yields bullwhip in the electronics supply chains, including semiconductor production, while Berry et al. (1994) undertook simulation modelling of a generic electronics industry supply chain to highlight the opportunities afforded by different supply chain reengineering strategies to mitigate bullwhip. However, their model did not explicitly represent the CODP and nonlinearities (e.g. shipment and inventory constraint, forbidden return) in the hybrid ATO system.

2.3.2. ATO system dynamics in the semiconductor industry

Overall, from a system dynamics perspective, very few studies focus on the ATO system in the semiconductor industry, apart from Gonçalves et al. (2005); Orcun, et al. (2006) and Orcun and Uzsoy (2011). Gonçalves et al. (2005) developed a system dynamics simulation model to explore how market sales and production decisions interact to create unwanted production and inventory variances in the Intel hybrid ATO supply chain. Using a system dynamics approach, Orcun et al. (2006) developed a capacitated semiconductor

Literature review

47

production model with load-dependent lead time, which overcomes the limitation of treating lead times as exogenous parameters independent of the decision variables that most linear dynamic models assume. The analysis suggested that nonlinear change at high capacity utilisation is consistent with insights from queuing models and industrial practices. Furthermore, Orçun and Uzsoy (2011) studied the dynamic behaviour of a simplified semiconductor supply chain system with two capacitated manufacturing echelons and one inventory echelon. They indicated that the dynamic properties of a supply chain system under optimisation-based planning models are qualitatively different from those operating under simple feedback policies system dynamics models.

Although these system dynamics simulations contribute to the representation of a real system by incorporating nonlinear components and complex structures, it is a trial-and-error approach that may hinder the system improvement process (Towill, 1982; Sarimveis et al., 2008). Despite the fact that semiconductor supply chains have suffered severely from the bullwhip effect (Chien et al., 2010; Terwiesch et al., 2005), limited research studies have explored the underlying system structures that cause the phenomenon. As a result, there is a need to consider analytical methods to understand the underlying mechanisms of bullwhip generation and propose corresponding mitigation approaches that are relevant for the semiconductor ATO supply chain.