Suppose you are the lucky winner of the lottery and suddenly in possession of 10 million Euro (or $, £ or Swiss Franc). The first thing you want to do is to buy a red Ferrari with yellow dots: the one you have been dreaming about and coveting for many years. So you go to your local Ferrari-shop and, with the friendly salesman’s assistance, you pick out your Ferrari – complete with all of the special gadgets and customizations you have always wanted. Like an extra built-in iPod and iPad for an additional 10,000 euro or a white goat skin knob on the gearshift for another 6,000 euro.
After some chit-chat with the salesman, the ultimate question inevitably will be: “When will the car be available?” Because you know the dream car you ordered will have to be customized, you are willing to accept a certain delay in the delivery. But the actual lead time hurts: 1.5 years to be exact, due to the limited availability of such a rare vehicle and to all of your special requests.
The salesman promises that the car will be delivered to your doorstep on Wednesday, March 12, 2014. So what is the worst that can happen on Tuesday, March 11, 2014? Probably a call from your local Ferrari dealer, telling you that the delivery of your car has been postponed. And alas, the new date cannot be revealed, yet.
Does this situation sound familiar to you? Well, maybe not with a red Ferrari covered with yellow dots, but for those of you in supply chain management who deal with other industrial goods, it will certainly ring a bell. The sales guy promises a delivery date but his production people apparently are unable to meet the quoted lead time. How is this possible in an age of lightning-fast cloud computing, S&OP processes, MRP-specialists and so on?
The heart of the problem is, of course, that sales promises and production has to deliver. And both have different objectives. But why is production unable to quote a realistic lead time?
Well, to be honest, I don’t know. Because lead times are tricky things. If we take a closer look at lead times, we realize that only a very small portion of lead time consists of production time. The production time of a BMW, Toyota or Ford is actually less than 1 day. The real production time of produced items in an industrial plant can be measured in minutes or maybe hours. But the lead times of these items in the same plant will be measured in days or probably in weeks.
This begs the question: What are those products doing most of the time? The answer is simple: waiting. Indeed, during most of the lead time products are waiting. Just like you at the supermarket. An efficient employee can scan something like 45 items per minute, a fraction of the time you actually spend in line. An answer from a call center most of the time takes less than a minute. But you will be in hold for muuuuch longer. So in order to control lead times we should control waiting times or queues – this, unfortunately, is not an easy task.
However, there are a few elements that can influence lead times to large extent and that managers should be aware of. First of all, the utilization rates of capacities. While financially-oriented managers want to utilize capacities at high rates, waiting times (and so lead times) increase exponentially. At a utilization rate of 90%, waiting times can be as much as 10 times the production time: at 95% it will be around 20 times! And to make matters worse, variations in waiting times can be just as high. So here is lesson # 1: Don’t utilize your capacities over 85%!
Now let’s investigate the second culprit: the priority rule. In typical commercial and social life, there seems only one acceptable priority rule: First Come First Served (or First in First out). From a logistical point of view, it can be proven that following this rule leads to mediocre or bad performance. And we see the consequences in reality! In a supermarket you have to queue behind customers who are buying tons of groceries while you only have one or two items. Some supermarkets employ a fast lane to address this situation, in this way introducing a new priority rule: Shortest Processing Time first (SPT). And yes, this rule makes overall performance much more efficient and smooth. Of course, one has to make some adjustments, otherwise customers with tons of groceries (and long processing times) would have to spend the entire night in the supermarket before being able to buy the goods they want. So here is lesson # 2: Use a SPT priority rule in order to minimize average lead times!
The third culprit is higher management. In many cases I have experienced that higher management wants to micromanage and control everything that is happening on the shop floor. Either by exerting influence in person or using a shop floor control system to do this. Apart from certain environments (such as process industry or mass assembly), this will actually be counter productive. Only the shop floor manager should decide on the final sequence of orders on the shop floor as he truly knows and understands the ins- and outs of equipment and personnel. Higher management should be responsible for the number of hours the shop floor should produce per week. The shop floor manager should decide on how to sequence these orders. The consequence of this procedure is that all jobs will be produced on time within a certain time frame. But higher management will not know which job will be finished first, and for many managers this sounds frightening. But this is exactly what every MRP system expects. So here is lesson #3 : Let the shop floor manager decide upon the final sequence of jobs!
If the Ferrari people understand the lessons above (or else quote a very pessimistic lead time), you might very well be able to zoom down the country roads in your new ride on Thursday March 13, 2014!
