In production environments with complex planning constraints, DDMRP inventory buffers are not sufficient to reduce the supply variability. This article explains why.
Part 2: Why DDMRP needs scheduling in a campaign-driven production environment
In the first part of this article series (click here for part III), we explained how DDMRP works and why companies must not neglect production planning constraints since focusing on inventory management only can lead to significant production inefficiencies.
The absence of production scheduling considering manufacturing constraints leads to capacity over- and underutilization, re-prioritization of orders affecting lead times and extended production time due to not optimized production sequence. As a result, short-term rescheduling often leads to a postponement of orders and, consequently, to an order backlog piling up at the end of the lead time. In addition, the risk of component shortages increases.
Figure 1: Planning situations driven by aggregated master plans lead to constant re-prioritization events operationally destroying flow with a supply chain out of sync
Even if buffer levels are used as prioritization criteria, it is unavoidable to have short-term sequence changes if scheduling activities are completely missing during the order creation activity. This noise within and at the end of the lead time is contradictory to the actual goal of DDMRP, the reduction of variability. As a consequence, supply variability would not be reduced and in the worst case the result is again a bimodal stock distribution when stocks are either too high or too low but rarely at the optimal level.
Another negative side effect is an aggregated, bucketed master production plan in the tactical horizon which tends to over- or underestimate capacity consumption. This leads to imprecise capacity plans and might require a rough overestimation of buffers to mitigate disaggregation effect in operational plans.
What we have experienced is that pure inventory buffer management is not completely hitting the goal for industries where production sequences really matter. The absence of any production constraint scheduling like changeover times may not lead to stock variability reduction. Shouldn’t the production planning variability be controllable, especially if sequence constraints are well known and predictable?
How companies can cope with these challenges will be explored in the third and last part of the blog post series. So stay tuned!