Traditional “MRP II” planning methods are struggling to meet the new challenges of the manufacturing supply chains, leading to shortages, overstocking and destruction.
Thousands of exception messages to move production or supplies forwards or backwards
Impossibility of making any kind of prioritisation
Which leads to a profusion of e-mails, Excel files and prioritisation meetings in “fire brigade” mode: a purchasing manager receives an average of 50 e-mails a day
Leading to shortages, overstocking and destruction
The logic of MRP is simple: backward planning of tasks with a breakdown of requirements using the bill of materials.
So what are the problems with this planning method?
1st example: Let's take the case of cosmetics, manufactured and packaged in our factories, then shipped by boat to our customers. For a customer requirement positioned at the end of April, the MRP retro planning logic plans the various stages of the value chain over time: transport in April, one month of security in March, packaging of the cream in February and production of the same cream in January.
In this first example, we will take the case of a delay in a supply order required for packaging.
Our supplier announces a delay in the delivery of our cases. Following classic MRP logic, an exception message appears on our order telling us that it is behind schedule. But is it really that critical? Are we going to put our end-customer out of business? Or are we simply consuming our safety stock? The MRP does not answer this question, it simply displays an alert. In this case, our supplier's delay can be fully absorbed by our safety stock, and will therefore have no impact on our end customer. The MRP alert is therefore not a priority. Problems highlighted here:
- Raw alerts withno notion of priorities
- The MRP is ON/OFF: each order movement triggers an alert, with no way of identifying its consequences or severity.
Consequences:
- INSTABILITY: Everything is constantly on the move
- LOSS OF PRIORITIES
2nd example: Now let's assume that the delay is a little longer.
This situation now leads us to a plan that is no longer realistic: given the date of arrival of my cases, I won't be able to pack my cream in February. If I don't reschedule my cream packaging, I won't be giving my customer any visibility on this delay. From his point of view, cream packaging is always scheduled for February, for delivery at the end of April. However, the suppliers in charge of the cases and jars are keeping a sense of urgency.
I therefore decide to adjust my production schedule to match the delivery of my cases, which will enable me to announce the delay to my customer: In this new situation, the MRP logic not only announces that the cases are on time, but also that the production of cream and the supply of jars can be postponed because they are too far ahead of the packaging order. The case supplier has lost sight of his priorities. For his part, the jar supplier thinks that demand has shifted, and has therefore also lost sight of customer priorities.
Problems highlighted here:
- MRP information is insufficient
- MRP has only one signal and forces an impossible trade-off between “building a realistic plan and losing customer priorities” AND “keeping priorities and not knowing when products will be delivered”.
Consequences:
- Loss of real customer demand
- Uninterpretable requirement signal
- Significant fluctuations
Conclusion
All these effects accumulate from one BOM level to another and amplify each other, making the MRP logic difficult to control, with impacts on service and inventory.
