Use this URL to cite or link to this record in EThOS:
Title: The Value of Information Sharing in a Multi-Echelon Supply Chain: a Simulation Study
Author: Syuhada, Maulana M
ISNI:       0000 0004 5349 3101
Awarding Body: Lancaster University
Current Institution: Lancaster University
Date of Award: 2014
Availability of Full Text:
Access from EThOS:
Information sharing has become the most recommended strategy in the literature to counter the Bullwhip Effect. The number of studies in this field has risen substantially since Lee et al. (1997a, 1997b) published their papers about the Bullwhip Effect. The studies have been predominantly carried out in two-echelon model of a supply chain using analytical approaches. Analytical studies are restricted by many assumptions for reasons of mathematical tractability, and often the model becomes unrealistic or too specific to allow useful conclusions to be drawn. To overcome this limitation, we use a simulation approach and model our supply chain as a multiechelon model where consumer demand is autocorrelated and the supply chain members do not know both demand process and its parameters, negative orders are not permitted and lead time is estimated using the realisation of actual lead time. An order-up-to policy is applied at each stage and demand is estimated using exponential smoothing. We observe that there is a strong relation between the service level and the lead time pattern. When the realised service level is not 100% the actual lead time will have an intermittent pattern with nonzero fixed base level. Whenever stock out happens the actual lead time will rise above the base level. The sporadic occurrences of non-base level lead time during stock out periods create a unique curve that resembles the shape of the stalagmites. Hence, we name it, "The Stalagmite Pattern". To the best of our knowledge, this distinct pattern has never been reported in the literature nor its implications investigated. In the Bullwhip Effect studies that model the supply chain as a multi-echelon model, the lead time parameters are very often underestimated, in which fixed lead time (FL T) is the most popular assumption, but no earlier research has examined this concern. The fixed lead time assumption only works well if the target service level is set close to 100%, otherwise the realised service level will be significantly below the target. The effect of the underestimated lead time is much worse at the lower target service levels. To overcome this problem we propose a model that uses the realisation of actual lead time to estimate the lead time parameters, called Smoothed Lead Time (SL T). Our simulation results show that the SL T that uses MEAN as the lead time estimate outperformed the FL T and performed consistently well for a wide range of target service level. This robust model is then used as our reference to find out the value of information sharing, which is the marginal benefit of using information sharing model relative to the traditional model where no information is being shared. Finally, we investigate the effect of demand autocorrelation, demand variance, and lead time on the value of information sharing to clarify some controversies about the benefit of information sharing reported in the literature. Our numerical analysis shows that information sharing reduces Bullwhip Effect substantially irrespective of the values of the autocorrelation coefficient and demand variability. It is most beneficial in reducing the Bullwhip Effect when the target service level is high, the autocorrelation coefficient is high, and demand variability is low. The benefit increases as lead time gets longer. However, the reduction in Bullwhip Effect is not always linear with the reduction in inventory level. Cutting lead time in half is a more effective method in reducing the inventory level than implementing information sharing. However, in term of bull whip reduction, both methods produce relatively the same results. The conflicting results in the literature regarding the benefit of information sharing in supply chain are likely due to the way the researchers configure the supply chain and set the assumptions.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available