Title:Mitigating Transient Bullwhip Effects Under Imperfect Demand Forecasts

Abstract:Motivated by how forecast errors exacerbate order fluctuations in supply chains, we use tools from robust control theory to characterize and compute the worst-case order fluctuation experienced by an individual supply chain vendor under bounded forecast errors and demand fluctuations. Building on existing discrete time, linear time-invariant (LTI) models of supply chains, we separately model forecast error and demand fluctuations as inputs to the inventory dynamics. We then define a transient Bullwhip measure to evaluate the vendor's worst-case order fluctuation and show that for bounded forecast errors and demand fluctuations, this measure is equivalent to the disturbance to control peak gain. To compute the controller that minimizes the worst-case peak gain, we formulate an optimization problem with bilinear matrix inequalities and show that solving this problem is equivalent to minimizing a quasi-convex function on a bounded domain. In contrast to the existing Bullwhip measure in literature, the transient Bullwhip measure has an explicit dependency on the forecast error and does not need the forecast to be a deterministic function of the demand history. This explicit dependency enables us to separately quantify the transient Bullwhip measure's sensitivity to forecast error and demand fluctuations. We empirically verify our model for vendors with non-zero perishable rates and order backlogging rates.