Demand Driven Planning – Element 4 of DDMRP

A DDMRP Recap:

There are 5 major elements to Demand Driven MRP (DDMRP):

  1. Strategic Inventory Positioning
  2. This page’s focus: Buffer Profile and Buffer Level determination
  3. Dynamic Buffers
  4. Demand Driven Planning
  5. Highly Visible and Collaborative Execution

Demand Driven Planning

Skip this intro if you don’t want context for this topic; it explains the causes of some problems we’re solving with Demand Driven MRP but if you just want the solution without context,  scroll down to  “The New Revolution.”

Few of us remember it (I lived through it) … but MRP was genuinely a REVOLUTION in the 70’s. This was already 20 years after MRP logic was initially codified; but it was in the 70’s when mini-computers made MRP available to a wide range of manufacturers, in contrast to the hugely expensive mainframe computers that were the only route for computerization until then, but out of reach for most companies).

The revolution was that MRP recognized dependency.

Before MRP, manufacturers might carry a stock of every part at every stage of manufacture, as well as materials and purchased parts, and manage the inventory of that stock (often using a Kardex card system) as if each part was entirely independent of every other, using the Re-Order Point technique. Inventories were high, stock-outs were common, both planned and actual lead times were very long (big batches) and on-time delivery and productivity were very poor.

MRP introduced the reality of DEPENDENT demand. If you needed one bicycle you needed 2 wheels; and if you had none in stock, you needed to make 2 wheels. For that you needed 72 spokes, and if you already had 40 spokes in stock and 20 were already on order then you needed to get 12 more.

When you added timing, and the concept of Master Production Scheduling and the refinements of forecast absorption,  you basically had a re-planning machine. Feed it the forecast, and the Master Schedule, and how much you had in stock for all your parts, & what purchase and work orders were already open; let it access the Bills of Material and part information such as lead times, minimum order quantities, and other order modifiers, and this terrific calculator would tell you what to buy, and how many, and when, and what to build and how many, and when. And, it’d give you reschedule recommendations for open orders to deal with changes in reality.

The good news is, this remains a great (re)calculator and today’s complex manufacturing environments with constantly changing realities calls for a darned good calculator.

The bad news is, it was never designed for today’s variability, volatility, and long (overseas) lead times.

From “nervous” to NERVOUS”

20 years ago we used to talk about doing battle with “nervous” MRP systems. In many environments small changes – in customer demand, or a stock adjustment, for example – would make the system “twitchy” and trigger a host of reschedules.

Well today – with variability, volatility, a demand for more responsiveness and agility and shorter lead times and better on-time delivery, while outsource supplies are creating longer lead times and more variability than ever before, and customer changes are more frequent and with less notice, and product life-cycles are shorter and new product introductions are more frequent and new materials are being introduced and new technologies being deployed … well, even writing these realities is exhausting.

But with all this in place, the MRP systems aren’t just “twitchy” any more, they’re in the throes of a perpetual nervous breakdown.

They’re churning out reschedule messages at a rate that Planners can’t hope to keep up with, while the fundamental data – the “what,” “how many,” ” when” – is now entirely unreliable because of the speed of change, the frequency of change, the Planners’ inability to respond to every single reschedule message in a timely manner, and the assumptions built-in to the systems.

As a result, companies are experiencing chronic shortages and inventory imbalances (too much of the wrong, not enough of the right) while productivity is being crippled, inventory totals are soaring, obsolescence being fed, expediting costs going through the roof and service levels are not even close to where they need to be.

This can be bad in even “regular” companies; it’s a nightmare in complex and challenging environments.

The Lean “Turn it off” … “No, keep it on” conflict

And, a special case exists when a company is attempting to implement a “Pull” system – most commonly Lean – and struggling to do so because of chronic shortages. One effect we see here is a heated conflict between he Lean advocates who consider MRP to be a “Push” system and demand it be turned off; and the Planners who (wisely) recognize that without MRP’s ability to recalculate needs, even with all its flaws, the company’s operations would be crippled.

The New Revolution

With Strategic Inventory Positioning, and the Dynamic Buffer concepts, Demand Driven MRP can bring:

  • Reduced inventory
  • Compressed lead times
  • Increased customer service
  • Increased stability
  • Less nervousness
  • Increased productivity
  • Fewer disruptions

To do this, DDMRP takes a step that at first glance can seem counter-intuitive – by building stock buffers in strategic locations, DDMRP DECOUPLES dependencies, which can almost seem like a backwards step.

But by decoupling at intelligent, strategic locations the “pros” – the bullet points above are only some of them – are enormous, and there are no negatives.

The issue that Lean advocates might jump on as being a “con” would be the perception that this creates more inventory, unnecessary inventory which is considered “waste” (muda) in the Lean movement. But in reality, these buffers ELIMINATE waste – they create substantial reductions in TOTAL inventory, while reducing disruptions to flow, reducing time in plant, reducing expediting expenses, and improving customer service levels.

Part-type designations for Demand Driven Planning

In Demand Driven Planning, parts are designated:

Replenished Parts – the parts strategically chosen to be managed as replenishment parts,  via a color coded buffer management approach. These parts will have dynamic buffers – they’ll increase when demand increases, for example, and decrease when demand decreases.

Replenish Override – also buffered and color-coded but NOT dynamic.

Min/Max – a variation of the classic min/max, with advantages of also being Dynamic; for readily available parts without the strategic value of Replenished parts.

Non-Buffered parts – these are classical MRP parts, not stocked, subject to regular MRP demand calculations. This will be the designation for the majority of parts.

Lead-time managed parts – non-stocked (so non-buffered) purchase parts that demand special attention. A time buffer is applied to these, to remind Planners to follow-up during the life of the order.

The reality of order spikes

Also, DDMRP needs to recognize the realities of there being “Order Spikes” for some parts, perhaps for some parts just for some customers. Ideally we’d like customers to recognize the value of a steady flow of parts to them, rather than occasional big “slugs” of work, but … for many customers, order spikes are a reality.

The characteristics of order spikes can be defined, and DDMRP constantly scans ahead to identify any spikes and ensure that the Replenishment system recognizes their size and timing, to ensure the Planning takes account of the spike.

Decoupling the Bill of Material

With Buffered parts, the Bill of Material explosion from Finished Goods to Raw Materials STOPS whenever it hits a buffered part; the stock Buffer decouples the Bill of Material.

The Zonal system is then the basis for replenishing the buffered parts, rather than the MRP-calculated dependent demand.

This reduces and dampens the nervousness, protecting against the variability cascading through the bill of material triggering reschedule messages in volumes that can never be dealt with.

The  stability this brings to some environments is stand-alone justification for Demand Driven MRP even without the advantages of improved availability, increased productivity, compressed lead times and right-sized (typically reduced) inventory.



Be Sociable, Share!