Please use this identifier to cite or link to this item:
Title: Optimal decision policy development for hybrid MTS-MTO supply chains
Authors: Wang, Fengyu
Keywords: DRNTU::Engineering::Industrial engineering::Operations research
DRNTU::Engineering::Industrial engineering::Supply chain
Issue Date: 2014
Source: Wang, F. (2015). Optimal decision policy development for hybrid MTS-MTO supply chains. Doctoral thesis, Nanyang Technological University, Singapore.
Abstract: The supply chain competition is increasingly characterized by high product variety, low volume and high service level. The challenge for the supply chain managers is to achieve responsiveness without sacrificing efficiency when dealing with high product mix. Relying on modular product design and manufacturing process standardization, more and more supply chains adopt a hybrid model that combines make-to-stock (MTS) and make-to-order (MTO) systems in sequence. Such hybrid systems can lower system cost by taking advantage of economies of scale of the MTS stage and satisfy the requirement of high product variety by taking advantage of flexibility of the MTO stage. Some researchers have found that the performance of hybrid supply chains suffers when the system capacity is constrained. This issue is not well addressed in the literature as most research is focused on strategy, and there are hardly any solutions available to develop effective operational decision policies, aligned with the overall strategy, that ensure system performance. The lack of support from operational decision policy for control of hybrid supply chains is a key research gap. This research proposes a joint admission-inventory control policy to bridge the gap that consists of a sequence of decisions made at the beginning of each (inventory) review period. The joint control policy makes the decisions to replenish the semi-finished module inventory at each review epoch, and sets the cap on the maximum number of orders that can be accepted during a period. However, the development of optimal joint control policy is a challenge. For a single period, the search space of the joint control policy is a multiple of reorder point, order up-to-level, and the number of orders that can be admitted in that period. Over multi-period planning horizon, the development of joint control policy faces the “curse of dimensionality” precluding an exhaustive search for the optimal policy in polynomial time. In addition, the performance comparison of candidate joint control policies is an issue as the optimization needs to satisfy the dual objectives of minimum cost and on-time delivery rate. The main objective of this research is to develop a more effective approach for the development of joint admission-inventory control policy. An innovative approach is proposed in this research to overcome the curse of dimensionality. By focusing on the codependency between admission control and inventory policies, a generic structure is formulated which can then be contextualized to develop specific joint control policies. The generic structure converts the optimization problem of joint control policy into identification of the optimal parameters of the generic structure, a combinatorial problem that can be solved by well-established methods. The advantage and effectiveness of the innovative approach is demonstrated and evaluated by employing Response Surface Methodology. By fitting first-order regression models to 250 local regions and fitting second-order response surface model to 15 local regions that are near optimum, the inputs are identified for the generic structure. The simulation results show that compared to the benchmark joint control policy developed using Bellman equation, the new joint control policy achieves a better service level that is closer to target on-time delivery. In general, the new policy derived from the generic structure performs very well. This research takes the hybrid supply chain research to next level; it is a necessary step to complement the state-of-the-art in strategy-based research. By establishing an alignment between the strategy and operational decision policy, this research benefits companies adopting postponement strategy that want to gain and sustain competitive advantage in time-based competition.
DOI: 10.32657/10356/65778
Fulltext Permission: open
Fulltext Availability: With Fulltext
Appears in Collections:MAE Theses

Files in This Item:
File Description SizeFormat 
Thesis Resubmission by Fengyu ver39 (20151112).pdfPhD dissertation2.77 MBAdobe PDFThumbnail

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.