Clariant's Matt Tichon asks: What is one of the primary goals of a chemical manufacturing company as it relates to their physical assets?

That is a relatively easy question to answer: reduce manufacturing and conversion costs and spread the fixed costs of production over the largest volume possible. We operate in a culture where talented engineers run Six Sigma projects to increase yields on individual lines, and increases in capacities and throughputs are celebrated - rightfully so within the context of manufacturing. The result is that large, efficient manufacturing plants exist, and if demand is questionable then we err on the side of production. After all, why would we not produce a bit more product? For sure that way if we experience an increase in demand, we have stocks for our customers, and we also ensure that the efficiency gains in manufacturing are realized.

What happens when the focus changes to working capital management, and we are trying to do what logisticians are trained to do: right product, right place, right time, and right quantity? That right quantity then takes precedent, as we want just enough in order to fulfill customer demands without tying up excess cash in stocks that are not needed for the market.

How to Determine the ‘Right Quantity'

Here is a task for you. I challenge you to ask almost anyone in your organization; from customer-facing sales people to upstream facing supply-chain planners how we can better determine the "right quantity." My guess is that you will hear a resounding cry for increasing our forecast accuracy. If we increase the forecast accuracy, then we will ensure that the "right quantity" is available.

In the quest for better forecast accuracy, we often turn to systems and solution providers, since they are the experts that hold the key to breakthrough performance. We invest hundreds of thousands of dollars, time, effort and organizational energy on increasing our forecast accuracy, often with the promise of finding the "silver bullet" that will be the answer to our woes. I can speak from experience that these systems, no matter how great they are at calculating a statistical forecast, mostly increase complexity within our functional organizations and end up not yielding the results that we expect or were promised.

This fact should remind us what the definition of insanity is: Trying the same thing, namely forecasting, with the expectation of achieving different results. We simply have to look for solutions that drive the results.

Getting Back To Basics

Think about how you manage your inventory in your own household. Do you forecast usage on items and ask that your local shopkeeper comes to your door with what you forecasted months ago? No - that would be silly! Most of us keep a list, either mental or written, for when we notice our stock running low, and we make a note of it so that next time we are at the store we can make a purchase and replenish our stocks. A rather a simple technique that works, too! Simplicity - that is what we need to get back to as we try and determine the right product and the right place at the right time in the right quantity.

As logisticians and supply-chain practitioners, wouldn't it be great to be able to satisfy our customers' expectations as well as the expectations of our corporation's shareholders with simply solutions that work?

Push to Pull

Fundamentally, we are all charged in one capacity or another with balancing supply with demand. What better way to do that than ensuring that we supply only what is demanded by the market? Let's look back at how the grocer replenishes the stocks. A "bin" - otherwise noted as "shelf space" - is set up for each product that the grocer is selling. The grocer operates using predictable replenishment cycle times, and monitors stocks on a real time basis via the point-of-sale system that records each individual item as it moves across the counter and is purchased by the customer. Once the stock level reaches a pre-determined point where stock will run low, yet not be depleted during the replenishment cycle time, a replenishment signal is sent to the supplier. The supplier responds, and the remaining stocks on the shelf continue to satisfy the customers' demands until the replenishment shipment arrives. This illustration is one of market pull based replenishment systems.
So let's look at how to practically apply this to our chemical segment as manufacturers.

Pulling Inventory from Upstream in the Supply Chain

Imagine a spreadsheet exercise where you look at the last year's worth of stock movements in your warehouse for each material in order to determine what your typical demands look like in any given week. Once you build this model, you could then look at your replenishment cycle times and set up a minimum inventory level, or reorder point, that is triggered when your stock reaches a point where you will just be able to further satisfy demand during your replenishment cycle time.

You could then use your materials requirements planning system to suggest and place a replenishment order once this level is reached, thus replenishing inventory only when the market has depleted it - then you pull the inventory back to the warehouse from the upstream in the supply chain. One of the other benefits of this system is that you no longer experience material requirements planning runs that are constantly adjusting replenishment signals for any given product as the system monitors pending orders that are routinely changed by your customers. Eliminating this tendency is possible because you only replenish once the physical inventory decreases.

Of course, you need to take this same concept and modeling analysis up the supply chain at all the various stocking points positioned between the point of manufacturing and the end customer's delivery point. If you take this approach, then the replenishment signal travels upstream, and the producing plant only produces when the market has already consumed. In order to operate on this principle, you fundamentally have to look at manufacturing planning in an entirely different paradigm than you may previously have done in the past. You may even now aim Six Sigma projects at increasing production flexibility and reducing campaign sizes; a radically different approach than that employed so often today.

Conclusions

Is pull-based, demand driven manufacturing the really the silver bullet we are looking for? No, it's not. Yet, it is a very powerful tool that we could and should deploy in a targeted way in order to balance supply and demand.
If you are able to achieve acceptable forecast rates then go for it; use your forecasting algorithms and classic material requirements planning runs. When forecast accuracy falls below an acceptable threshold a specific material, you may just want to try getting back to the basics; and if you do you may be pleasantly surprised to see how simple balancing supply and demand may be!

Most importantly supply chain managers can deploy this today, without having to compile and justify a capital acquisition request or rely on what are often constrained internal information technology resources. After all, isn't that what our organizations expect of us, to get results with the resources that we have at our fingertips?!