Demand-Driven MRP – Part II: Why is DDMRP needed?
This post is part of a series about Demand-Driven MRP (DDMRP). While the first part was an introduction to the basics of DDMRP, in this second part I want to discuss why such a solution is important nowadays. In the next posts, after we have all agreed on both the advantages and the shortcomings of traditional MRP, we’ll details the steps in the DDMRP solution.
Good old MRP
MRP has been around for 50+ years now, and this post-war engine is still the core of ERP systems. Many would argue it is the worst form of materials planning, except for all the others. Others, including some die-hard anti-IT lean advocates, do not consider their job done until MRP is switched off. Most in the industry, I think, have a sense that they can’t live without it, and they can’t live with it.
The logic behind MRP is pretty simple: based on your master production schedule (MPS), you explode your bill of materials, netting and time-phasing the requirements at each intermediate level, and end up with the raw materials requirements. In essence, it is the perfect JIT system, as its goal is zero inventory at each level.
VUCA – Volatility, Uncertainty, Complexity and Ambiguity
The first problem, of course, is how you come up with your MPS in the first place. This was probably easier back in the day when plants had insufficient capacity for the demand out there, and you planned your schedule according to your own best interests; nowadays, on the contrary, the world has spare capacity and the decision of what, how much and when to produce is more critical than ever.
Ideally your MPS would be based on existing customer requirements. Unfortunately, life is a *****, of course, and that is seldom possible even if you work in an MTO environment. Every manufacturer seems to be locked in the worst part of the supply chain: your suppliers have longer and longer lead times (think global sourcing), while your customers have shorter tolerance times. And not only the supply chain, but the products themselves are getting more complex and (mass) customized.
All these mean that you have to build your MPS based on more and more unreliable forecasts. In a VUCA world, which is the new normal and if anything will only get worst, forecasting is getting more and more difficult. To make things worse, someone high in the organization has read a book that says you can get by with zero inventories.
In MRP everything is interconnected, and this is at the same time its greatest strength and the source of most planning headaches. If you have been part of an MRP implementation in the past, you’ll probably have seen how everyone insists that planners always follow MRP messages and plan by exception, only to see them ignore these messages when you turn your back. Why? Simply because there are always too many of these messages and they are always changing, and people quickly stop trying to chase the wind.
MRP nervousness, something like an internal bullwhip effect if you will, is not a new concept: Orlicky’s original MRP book is from 1975, and the first paper I know about nervousness, by Daniel Steele, was published in… 1975! MRP books usually illustrate this with an example where a customer order cancelation actually leads to a previously feasible plan to become unfeasible.
The fact that most planners feel they need MRP (how else to explode finished product requirements into raw material requirements?) but still use their own Excel-based tools is a clear sign that there is definitely room for improvement.
The power of decoupling
As the next posts in these series will show, DDMRP attacks the problem by decoupling the bill of materials at strategic points. Also, by using a dynamically calculated average usage, which can be manually tweaked based on the planner’s knowledge, the system becomes less reliant on forecasts. From push and promote, to position and pull.