DDMRP stands for “Demand Driven Material Requirements Planning” and is a method of material requirements planning that is primarily based on actual demand. DDMRP combines traditional material requirements planning (MRP) with the pull and lean approaches to ensure better responsiveness and flexibility in the supply chain.
One feature of DDMRP is so-called “buffer zones” or stock buffers, which serve to balance out fluctuations in demand and supply and thus enable a continuous flow of materials. They are placed at selected points in the value chain and not at every inventory level, as is usually the case with planned replenishment and reorder point control. Criteria for positioning the inventory buffer are long delivery times, critical components or high fluctuations in demand.
Similar to a traffic light kanban, the stock buffers are adjusted according to a color-coded traffic light system. As long as the stock is in the green area, no reorders need to be carried out; if the stock is in the yellow area, a reorder must be triggered. If the stock slips into the red zone, there is an urgent need for action. DDMRP claims to avoid excessive inventories and at the same time improve product availability by looking at actual demand signals.
Our tip:In our experience, DDMRP is an approach that attempts to combine planning-based replenishment and reorder point control without offering any real substantial advantage. The correct idea of not having to use safety stocks at every inventory level can also be applied to classic replenishment procedures.
With DDMRP, stock buffers are dimensioned simply and pragmatically using flat-rate calculation mechanisms that incorporate a lot of gut feeling on the part of the person calculating the buffers. This can lead to certain buffers being far oversized and others far undersized. This can even lead to increased inventories on the one hand and reduced delivery capacity on the other.
Precise and coordinated coordination of the stock buffers is therefore of crucial importance when using DDMRP. You should simulate in detail and with time-dynamic data whether the buffers really achieve the effect you expect. A purely static view based on average values is not sufficient.
