Do We Need a Truck Preparation Area?
Dear Gemba Coach,
As logistics manager, I disagree with our lean expert who insists we should have “truck preparation areas” in the shipping hall. It adds more handling and goes against the lean principle of touching every container the least possible. As we have many trucks leaving per shift, it would also require a lot of space at the shipping docks and I will not build an extension for lean requirements. Am I missing something? Is our lean expert just following the dogma?
Building a warehouse extension for lean reasons wouldn’t be very lean, I agree, and a definite no-no. Actually, lean should help you get rid of warehouses, not the other way around. Nonetheless, truck preparation areas are indeed a classic lean trick and the first place I personally “go and see” when I do a plant visit. Before we debate about how valid or relevant the solution is, let’s go to the Gemba and try to clarify the problem we’re trying to solve.
A generic four-step improvement program for any value stream can be described as:
1. Protect the customer
2. Control lead-time
3. Reduce lead-time
4. Reduce costs
From a logistics point of view, protecting the customer means having exactly the right containers in the right truck. Typically, the customer might be either an assembler or a distributor. In any case, it should be possible to establish a precise manifest of which containers should be placed in which truck in order to deliver on time in full. In a perfect world, all parts should be available as and when necessary, but in reality we all know that many mishaps can happen, and when shipping operators go and look for the parts in the finished goods warehouse or the production shop stock, one container can easily be missing.
If a container of finished goods is missing from the stores and the truck is already there, waiting to be filled up, there is no time to react. The truck will leave with a missing container. The only practical way to make sure the truck will be complete in full is to assemble it ahead of time to see whether all parts are there. One can argue that you can make the same check through the computer software, but this assumes the software is perfectly up-to-date in terms of what exactly is in the warehouse. Experience shows there are frequent discrepancies between what the computer believes is in the warehouse and what parts can actually be found there. The only way to be really sure is to touch the parts themselves.
By deciding the truck preparation area (TPA) corresponding to one truck should be completed some time before the truck’s departure, it’s possible to let yourself the time to respond if one container is missing. Upon discovering that a container is missing, if the process is flexible enough, a couple of hours are enough to stop the current production, run enough parts of the missing reference to fill a container, and get those to the truck on time – thereby protecting the customer and delivering on time in full.
But what about the argument that this creates an unnecessary handling Muda? Well, there is a second reason for having truck preparation areas. Beyond protecting the customer, the second step of lean improvement is controlling the lead-time. If the parts are brought directly from either the finished goods warehouse to the truck or from the production cell’s shop stock to the truck there will be unavoidable variation in lead-time. In the first case because we don’t control the time it takes to find the parts in the finished goods warehouse, nor do we control how they get there. In the second case because the parts would sit in the shop stock and then be drawn in one go according to the truck’s manifest. Both cases will create a variation-riddled situation that will increase the lead-time of moving the parts from production to the truck and then on of production scheduling.
The TPA allows us to break up the anticipated truckload into a regular pitch in order to withdraw from the previous point of stock steadily, thus controlling that part of the lead-time. Whereas a real truck (which needs to be loaded and go) would have to be filled all at once, a virtual truck represented as a TPA can be filled incrementally through the shift at the same pace as the customer uses the containers on their line or shops. This is an ideal, in many cases, as the customer plant’s pace and the suppliers are not exactly the same, so there’s some fooling about with filling some TPAs in advance in order to have a steady pace of building up the truck loads as well as let the customer pick up at their leisure.
Create a Steady Pace
Pulling on the line regularly in order to fill up the truck preparation area at an even pitch has the added benefit of giving a steady pace of production to the cell, and creating the kind of tension conducive to lean work: if the line is late, it will show up immediately and the supervisors can react immediately and solve the issues at the root of the delay. Not having a truck preparation area would impose a stop-and-go rhythm on the line (mura) which would make kaizen very difficult.
In case of working for distribution centers it’s often hard to know what the truck load will look like until the orders come in through the day. In this case, a finished goods warehouse stock is necessary, but the same argument applies. Building up the truck as the orders come in at an even pitch will keep the finished god stock operating as a buffer rather than real inventory, which will help in controlling the lead-time for the shipping part of the process.
In any kanban system, controlling the upstream part of the lead-time is essential to heijunka and smooth production as the information is transmitted down stream through the kanban signals. In this sense, the truck preparation area is a necessary tool of the kanban system. Without a way to pace the pull of finished goods, the lines would have to deal with unbearable variation in lead-time, which would make smooth production and kaizen very difficult.
Key to Cutting Lead Time
Come on, you will say – live in the real world: how can we control our lead-time when customer demand itself has so much variation? Here again there are a number of tricks, from talking to the customer, to creating buffer “pool” stocks to handle the variation: separate inventory places where the difference between the leveled “perfect” customer demand and the real customer demand is kept. Whereas the Truck Preparation Area should hold the exact customer demand, the leveling board giving instructions to withdraw on production should be the image of a “perfect” customer – a completely leveled customer. The discrepancy between this “perfect” customer and the real demand is in fact lead-time variation that we can place in a dedicated area and first control and then reduce by improving our communication with our customers.
The truck preparation area is also the key to reducing lead-time once this is in control. In many cases it’s hard to think of ways to reduce lead-time because that would involve more customer trucks coming to pick up – and then theses trucks wouldn’t be as full, which would create waste by moving air rather than parts. But this is also an opportunity for kaizen. If the customer trucks withdraw several parts from the same facility, are we certain each truck picks up a little of every type of parts, or do they withdraw, as usual, full truck loads of one type? Even pallets made of storage units need not necessarily hold only the same product. The idea is to divide and mix the truck loads so that every truck lifts a bit of everything rather than have one truck full of one product and the next full of another. Doing so requires a pallet building stage, which, in practice often necessitates a further handling point before the truck preparation area (which consolidates pallets).
How is any of this ever going to reduce our costs? All we see so far is a multiplication of storage points and handling: more Muda, not less! Not really, and that’s the magic of lean. Firstly, pulling steadily from the line can be done with a regular train as opposed to forklifts. This is a massive cost improvement because, if you stand in any warehouse, you’ll see that for every forklift running loaded you can find at least one coming back empty and another sitting idle. In every plant where we’ve created a steady pull of parts with a regular trained crew we’ve decreased transportation equipment cost dramatically.
Secondly, labor costs will also decrease sharply as less people are needed to drive the forklifts and run after the parts This will also avoid the perennial waiting times and traffic jams due to the fact that, thanks to Murphy’s law, all forklifts always have to be at the same intersection at the same time, and will also create a safer working environment by separating man and machine sections. A few forklifts might still be necessary to load the pallets on the trucks, but these can be confined to a “no walk” area. If the system is pushed to its logical conclusion and the pallets are ready to be loaded, the truck driver himself can open his truck, jump on a forklift, take out an empty container, and replace it by a full container and load the truck himself rather than wait idly at the wheel.
Magic Wand Wanted
In terms of shipping area space, yes, it is true that there is an awkward moment when you first establish the TPAs – it usually involves moving some inventory away somewhere else. But as soon as the truck preparation areas are properly used it becomes obvious that less inventory is needed in order to fulfill customer orders. Progressively controlling and reducing lead-time will release more and more storage locations until all inventory can be held in virtual trucks prepared ahead of time rather than a finished goods storage warehouse – the end result is likely to be about half as small as when you’ve started, if not two-thirds smaller. So draw a deep breath and look for ways of creating the first TPAs rather than judge ahead of time whether it will work or not.
Lean techniques rarely come with a magic wand – at the best bring a lot of head scratching and honest sweat. There are rarely comfortable either. In lean tradition, most new tools can be greeted by an expression of “Oh No!” as a nod to Taiichi Ohno, the inventor of many of these tools. Lean tools are support for kaizen, not the other way around. The upshot is that it’s hard to prove the benefit of any tool without actually working with it. Any cost-benefit assumes “all things being equal”, but the very nature of the lean tools is change, so all things are never equal and the benefits come from unexpected places, challenging assumptions deeply imbedded in financial accounting.
I am aware that you heard exactly the same thing from your in-house lean expert and that you probably don’t feel anymore reassured by having read this. Unfortunately, it’s in the nature of lean that learning by doing can only be achieved by, well, doing. If nothing else, the first step rationale for the truck preparation area is to make sure we’ve got the truck’s contents ready before the truck gets there, which will secure both the delivery and the loading process (it’s quicker to load than if we’ve got to go and fish for parts in the warehouse). Having passed through that first door, you’ll find out that controlling the lead-time at the customer truck level is the key to the stability of the entire pull system, which, in it turn is the key to dramatically reducing inventory and improving radically the cash flow – so all in all, it’s worth some hard thinking to figure out ways to make Truck Preparation work, rather than argue that they’re wasted time and effort.
How can I make sure my teams do kaizen the right way?
Dear Gemba Coach,
How can I make sure my teams do kaizen the right way?
How do you apply takt time to service work?
Dear Gemba Coach,
How do you apply takt time in fields like services where customer demand is not known?
Review: Designing the Future
In his review of the new book Designing The Future, Michael Ballé points out that it “makes clear the central lean concept in product development: distinguishing what is fixed and what is flexible in new product design.”