Faculté des sciences économiques et sociales

Supply chain performance : the impact of interactions between flexibility enablers and uncertainty

Nieto, Yvan ; Reiner, Gerald (Dir.)

Thèse de doctorat : Université de Neuchâtel, 2011 ; 2237.

This doctoral dissertation in the field of Supply Chain Management focuses on the dynamic interactions between uncertainty, flexibility, and performance in the supply chain. In particular, it provides evidence that the improvements related to uncertainty reduction practices are modulated by the flexibility enablers engaged by the companies involved in the supply network. Consequently, it... Plus

Ajouter à la liste personnelle
    Summary
    This doctoral dissertation in the field of Supply Chain Management focuses on the dynamic interactions between uncertainty, flexibility, and performance in the supply chain. In particular, it provides evidence that the improvements related to uncertainty reduction practices are modulated by the flexibility enablers engaged by the companies involved in the supply network. Consequently, it highlights the need for system-wide evaluation that captures the dynamics of the specific operational characteristics of a network. Further, this dissertation explores the issue of the interaction between uncertainty and flexibility in model-based supply chain evaluation. The results denote that neglecting these relationships might lead to mistaken estimation of the supply chain performance.
    The impact of both uncertainty and flexibility on supply chain variability leads to multiple dynamic interactions, which creates complex problems at the time of assessing the impact of supply chain improvements on the performance of the system. Indeed, if general supply chain behavior has been identified, the impact of the interactions between specific operational settings on the system dynamics remains largely unknown. In particular, serious gaps remain in the understanding of the dependencies between flexibility enablers. This understanding is important because of its direct implications on the dynamic of the system. Interactions between flexibility enablers are assumed to influence the general ability of the system to demonstrate flexibility and therefore directly affect the behavior of the system. This knowledge is therefore required in order to better understand the mechanisms underlying supply chain performance, and then to enable model to capture better the dynamics of the systems.
    The core of this dissertation is constituted by three articles that can be read separately. All of these studies are related to the interaction between uncertainty and flexibility with regard to supply chain performance; however, the detailed research questions considered in each study are distinct.
    The first study investigates the interaction between forecasting and flexibility enablers with regard to managing demand. It is assumed that superior knowledge regarding demand influences the production process as well and, consequently, the flexibility enablers' effect on company performance. This analysis evidences that the impact of forecasting on performance is due to the mediating effect of flexibility enablers. In particular, it shows that the relationship between forecasting and customer satisfaction is mainly due to process flow management, while the relationship with cost efficiency is mainly due to layout. Therefore, this study not only provided evidence of the existing relationships between forecasting and performance, but it provided as well some insights on the causality of this relationship. In particular, the study showed that the availability of additional information helps the actors involved in a specific process to make prompt decisions and align different units that manage each separate part of the production process. The performance improvements (i.e., increased cost efficiency as well as effectiveness) related to improved forecasting can then be magnified by process or layout modifications.
    The second study examines the real-life situation of a supply chain where the flexibility of the production facility is directly influenced by the level of uncertainty in demand. In this setting, better demand information (i.e., uncertainty reduction) allows for wiser decision making regarding capacity adjustments and reduces the trade-off between volume flexibility and production efficiency. The simulation results shows that, for the specific settings of the study, the benefits from the uncertainty reduction provided by advance demand information are strongly influenced by the production and inventory management constraints engaged in the supply network. In particular, the value of advance demand information is magnified by the realistic situation where uncertainty reduction enables the manufacturer to improve the alignment of its production capacity with the forthcoming demand. Also, the value of the advance demand information is dependent on the policy regulating the interaction between the distribution center and the manufacturer. These interactions combine to determine the ability of the supply chain to satisfy demand in an effective and efficient manner and are therefore of critical importance to support decision making.
    The third study considers the implications of integrating the non-constant variability of the demand variance with periodic reviews of up-to-level inventory policy. Such variability is traditionally observed in seasonal demand patterns where the demand variance is generally correlated to the average demand. This research work demonstrates the potential trade-off existing between safety stock accuracy and safety stock variability in inventory control with seasonal demand. Specifically, it highlights how the improvement achieved in adapting the safety stock to the fluctuation of the variability of the perceived uncertainty (i.e., the non-constant variability of the forecast error) is balanced by the resulting increase in replenishment order variability. Further, this study demonstrates that this dynamic dependency can only be captured if the relationship between the capacitated production system and the inventory system is explicitly modeled (i.e., the lead time is endogenous).
    In general, the above-mentioned contributions highlight the need for a better understanding of the relationships between uncertainty, flexibility, and performance at the supply chain level, integrating in the performance measurement both efficiency and effectiveness. Further, it provides evidence that valuable estimation of performance based on supply chain modeling requires integrating the detailed interaction existing between the specific operational characteristics of the network under study. From a managerial point of view, these findings stressed the danger inherent to decision making supported by over-simplified supply chain models and propose, through an example, a methodology to tackle this issue.