Looking at the current development of the pandemic, everything indicates that we are in a vicious circle. After a wave of successive easing of restrictions, governments in many countries are considering revisiting restrictions on freedoms. No one knows how deep the restrictions will be, or when they will be implemented. What is certain, however, is that they will cause further perturbations in supply and demand behavior.
Is it possible to chart the course in which our supply chain optimization decisions should be headed before we affect the rocks ?
A solution that is certainly worth considering is to build an early warning system to catch situations of potential risk of out-of-stock and overstock situations in advance. To build such a solution, on the one hand, we need a tool for generating accurate demand forecasts, i.e. One that will allow models to quickly adapt to changes in market behavior. On the other hand, tools for frequent analysis of future inventory projections.
The accuracy of the forecasts can be improved by using several elements. The first is a machine learning mechanism that allows the simulation of many different forecasting models for each SKU and the selection of the one that best matches the demand pattern at a given moment. Since demand these days can change rapidly, such a model should be selected each time new sales data is entered, such as updated once a week. The second element is to clean up historical data from Outliers and reduce their impact on current demand in the forecasting models used. The third is to take into account the forecasting impact of so-called “calendar days” that have a significant impact on demand patterns.
In order to carry out future inventory projections, it is necessary to know both the demand forecasts as well as the real constraints in the supply chain and the business objectives (most often expressed in terms of required availability of the so-called Service Level). Real constraints in the supply chain can be captured by, among other things, tracking the availability of our suppliers or their Service Level (deviations in terms of quantity as well as time).
Accurate forecasts will allow us to simulate the consumption of inventory by demand, and knowledge of real constraints in the supply chain will allow us to calculate the timing of orders and the optimum size to reflect our business goal, i.e., to make the best possible use of inventory. required product availability.
The main feature of the model based on the so-called. “ordering point,” also known as an information-level ordering system or continuous review , is a condition […]
According to Gartner, the level of automation in supply chain processes will double in the next five years. At the same time, global spending on IIoT […]