Cognitive Mapping, System Dynamics and the Bullwhip Effect

This project deals with the problem of the Bullwhip effect, i.e. that of magnifying inventories as we move backwards in a supply chain. This effect has been highlighted in the academic literature for many decades and, admittedly, it has attracted a tremendous amount of scientific research. However, all the relevant projects have focused on only a limited number of the contributory factors, with no integrative model having yet been established. Despite the importance of such work for developing our understanding of the behaviour of supply chains, the problem still prevails in industrial applications. The fragmented approach to the relevant problem solving may well be explained in terms of the considerable associated complexity. That is to say, the contributory factors (e.g. demand signal processing, rationing/shortage gaming, order batching and price fluctuations) have to be viewed as interrelated components (as they are in practice) rather than stand-alone issues of a wider formulation, and this obviously increases the theoretical complexity of the problem. Analytical solutions cannot be developed unless some of the factors that contribute to the effect are isolated and tackled separately. In addition, and in order to facilitate the mathematical treatment of the problem, many assumptions need to be made, the validity (or universal applicability) of which has also been questioned. It is viewed as imperative to introduce novel holistic approaches in order to solve complex supply chain/ inventory problems, such as the Bullwhip effect. These complex problems demonstrate the industrial importance of inventory management and the considerable benefit that their solution may offer to modern organizations. Such a solution necessitates an inter-disciplinary approach to problem formulation and modelling and this is what out proposal introduces. There is evidence to suggest that the integration of System Dynamics models and cognitive maps constitutes a very promising approach to group model building and is particularly appropriate for supply chain models with significant behavioural components. The aim of this dedicated 3-month project is to re-conceptualise the complex problem of the Bullwhip effect through the integration of Cognitive Mapping and System Dynamics (stock and flow) modelling. Such an approach allows for the consideration of behavioural factors in addition to exploring interactions. Two such factors are highlighted in our research: i) judgemental changes to forecasts and ii) judgemental changes to replenishment orders. Both have been shown to prevail in industrial practices and they are appropriately incorporated in our models. In addition to researching the extent to which such models represent reality, we also intend to explore their potential to act as a training tool for those new to judgemental forecasting and, in the longer term, to change company policies.