Scanning the Internet for External Cloud Exposures via SSL Certs
Best of the Best Sales and Operations Planning Conference Using S&OP in Companies
1.
2.
3.
4.
5.
6.
7.
8.
9.
10. Stock Amount In this strategy there is no correct level of inventory
11. BOM Shape helps dictate strategy In this ‘shape’ a few end items are made from many parts, materials, and components. An example would be an equipment manufacturer such as automobiles, lawn mowers, capital equipment, etc. This shape lends itself to making to stock. Levels # of Parts
12. How Lean fits in: Make to Stock Supermarket Here Customer Lead Time Replenish Finished Goods on pull signal Or create mix leveled Heijunka schedule and push to finished goods Pull or Push to finished goods
21. BOM Shape In this ‘shape’ many end items are made from a few raw materials. An example would be an injection molding company This shape suggests a make to order strategy.
30. BOM Shape In this ‘shape’ many end items are made out of a few sub assemblies or intermediates that are made out of many unique raw materials and parts. An example of this would be a pharmaceutical manufacturer with many package types from standard product (aspirin), personal computers, packaged chemicals This shape lends itself to a postponement strategy.
31. How Lean fits in : Postponement Supermarket Here Customer Lead Time Pull Components based on usage Assemble to order Backlog management here Stock management here
This presentation is the work of Bill Kerber, and although you are welcome to use it for good, please give credit where credit is due. Thanks.
While Lean companies certainly must do long term planning just like any other company, the literature rarely discusses how this is done, instead focusing on the shorter term activities such as Heijunka. The slide attempts to make it clear that the Executive S & OP process is a natural fit in a Lean environment. The rest of the presentation will deal with some of the issues that should be addressed as you implement Lean with Executive S & OP.
Takt time is an important concept in Lean. Takt time is used to set the rate of sales for the value stream. The word Takt can be interpreted to mean “beat”. The question that we need to think about in conjunction with Executive S & OP is whether the traditional use of takt time will serve for all of the possible production strategies supported.
Effective working time is the amount of total time minus paid breaks and lunch. The contention that I am making here is that the only environment (or production strategy) where this definition really makes sense is if the value stream is employing a chase strategy, ie: sales and production are designed to be the same in the interval. For many environments, however, this strategy is not the least wasteful one, for instance where a seasonal business pursues a level or hybrid strategy, and therefore has many intervals where the production and the sales rate will not be equal.
Pitch is the preferred method of measuring to see if a value stream is producing at the takt rate. If I am going to measure if the value stream is producing at the correct rate, it would be better to measure at the agreed upon production rate in all cases, as opposed to the sales rate which will be meaningful to the factory floor only in the case of a chase strategy. An easy to understand example is the following: if inventory is higher than desired, the only way it will reduce is to produce less product than is being sold. If takt rate and the value stream are configured and staffed to produce to the demand rate, inventory will stay constant. Only by configuring and staffing to the Executive S & OP production rate can we change inventory (or backlog) levels.
We could also correctly call these supply strategies, since companies must decide the approach to the market even if they do not do a formal Executive S & OP process.
This would be a good place to discuss the three “M”s of waste in Lean, namely Muda, Mura and Muri. There has been lots of focus on Muda in Lean so far. APICS and IBF members are in a unique position to help with eliminating Mura (variation) in the value streams. Selecting the correct supply strategy is important in this effort.
It is necessary to understand a bit about the Toyota environment to make sense of what is written about how they do this versus what I am presenting here.
Inventory is the most common method used to buffer value streams. The usual approach to this is to implement a Kanban system for the finished goods products,. This would mean that all items are in the finished goods inventory and would be shipped on demand, at which point the value stream would receive a signal to simply try to replace what was shipped. What this does for the purposes of this book, however, is to simply pass the demand variation of the marketplace into the pacemaker process of the value stream. The key concept of using inventory as a buffer should be to provide level work for the value stream. The result of this is that having inventory should not be the goal; it is instead the result of the production strategy. What this means is that the presence of inventory allows the value stream to produce a constant amount of work even when the demand from the customer is erratic. Because the demand will vary and the production rate will not, this means that inventory will vary in the opposite direction from demand, i.e. when demand is high, inventory will go down, and vice versa.
In this ‘shape’ a few end items are made from many parts, materials, and components. In this circumstance, the demand for lower level components will tend to be dependent on both the volume and mix variability for end items. Because the demand for the lower level components will tend to be very mix sensitive with this type of shape, Toyota has chosen to completely level the mix within the factory, and allow the finished goods to absorb variation in both mix and volume. This strategy calls for the possibility of all finished items to be in the finished goods inventory. They may not necessarily be in finished goods at any one time, but because of the way the strategy works, any one of the possibilities for finished goods combinations may end up in inventory. In practice, the ones that sell less than the dealer forecast are the ones that will end up in inventory.
Using the word push here is intentional and should spur some discussion. Toyota does not, by strict definition, push product into the value stream Everything produced in the plants are made to an order. However, it is not really a demand pull through the value stream, since the order from the dealer is based on forecast of future demand, not on actual demand. It would be my contention that in the face of imperfect markets, ie: the customers buy different quantities and mixes in each interval, there will need to be at least some items pushed to obtain a level schedule in a make to stock environment. This slide places the strategy into the context of Lean. The supermarket is placed at the narrowest point in the BOM shape, namely at the finished goods level. The production of product can then be based on a pull from the supermarket, (with some allowances for leveling). The customer only has to wait for the product to be shipped, so lead time to the market is kept to a minimum. The lead time through the value stream to replenish the supermarket helps to determine the inventory level, so speed counts in this strategy.
We should make an upfront distinction between selling a level mix and producing to a level mix . Toyota produces to a level mix but clearly does not sell a level mix to the public. What actually happens is that they produce to a level set of orders from the dealers, who have an inventory which absorbs the difference between what is sold and what has been produced in any period. The activity at the plant is to take the dealers orders, and parse them up into equal segments, so that each model and option shows up in the schedule at the same rate as it has been order, for instance half of the cars ordered have DVD players, so every other car in the schedule will have a DVD. Volume variability is only concerned with whether the sales can be covered by the capacity within the plant on average basis . In other words if we average selling 5,000lb. of product each week , how far away from the 5,000lb. are the actual sales during any week .
This is an example of how mix is leveled. At Toyota, this is against the car orders placed by the dealers. In other environments, it may represent the forecasted amount for each option. The key is that production makes the average amount of all products, without regard to the actual sales to the genereal public as they are realized. Next month, if the sales mix has changed, then the sequence will be reset to reflect this change.
There are many advantages to completely leveling the mix in a value stream. Most of these revolve around the fact that most long term planning involves some sort of averages for the product required in the value stream, from the amount of equipment needed to the manpower employed to the size of supermarkets. Because leveling the mix means producing the average, the decisions made in the long term planning will generally work as designed.
If a company does decide to level the mix in a high mix variability value stream, there will probably be some poor results. This list is by no means exhaustive, but these are still serious considerations for management.
Because in some value stream the high mix of products does not allow for leveling of the entire mix of products without a significant amount of waste, the strategy must somehow level work into the value stream without the waste accompanying a full supermarket of slow moving items. This can usually be done by keeping a supermarket of only high volume, frequently ordered items, and treating the rest of the items as make to order items. The high volume items have several advantages as a buffer: They are less likely to become obsolete, and, because they are frequently ordered, it is easier to adjust the inventory level on them. This strategy requires the separation of the supermarket items from the rest, with the following rules usually in place: Load the Heijunka box with the make to order items first, then fill the remaining capacity with the supermarket items. Because not all of the items will be produced in each interval, the interval needs to be revised to be the required items to build in a period, based on what is actually sold in the period, not all of the products in the family.
Because we will not be producing the average product mix in any period, we are vulnerable to the swings in work content caused by the various product mixes that are needed. In other words, we may have “Floating bottlenecks” in the shop. This means that it is not sufficient to simply level the work at the pacemaker, but to also check to make sure that all of the other resources can produce the parts necessary to make the product mix. This requires a bit more sophistication in the capacity planning portion of the value stream. For the same reason, it is necessary to have more fully cross trained operators when only leveling volume, since to some extent it should be expected that operators will need to move around to where the work is for any particular product mix.
This is a chart representing the strategy of buffering a value stream with a variable backlog. The slide shows that backlog will vary with the sales volume, ie: if sales are higher than production, backlog will increase, and if sales are less than production than the backlog will get smaller. Backlog in this environment will equate to the customer lead time.
In this ‘shape’ many end items are made from a few raw materials. An example would be an injection molding company, making many different end items, all from a common resin and a few additives. In this circumstance, the demand for lower level raw material will tend to have little difference from the volume variability for end items and would be unaffected by mix. In such an environment, it would be possible to stock the raw materials based on the volume leveled plan, then produce to the customer order.
The key to producing to customer order is to ensure that the production lead time is less than the market lead time. Lean obviously plays a major role here, as one of the usual results of a lean value stream implementation is shortened lead times due to flow.
Having understood that the manufacturing lead time must be less than the market lead time, this slide suggests that it must be even less than expected. This is because the period of time from when the order is received until it is started in production must also be included as part of the lead time that the customer will experience. The expected amount of the backlog will depend on the amount of volume variation experienced by the value stream.
Not all markets will accept variable lead times. In this case, we must make the promised dates on customer orders based on a fixed lead time, and we must make the combination of the un-started backlog plus the production lead time less than the market lead time. These two segments of the lead time must be monitored and managed, because if the total grows to be longer than the market lead time overtime must be worked. In other environments, we can promise variable lead times, in which case the promise date can be made based on the amount of work already sold versus the capability to produce the work. This must be done at the aggregate value stream level, not via something based at the item level such as ATP. But that is another topic for another day…
A chase strategy in theory has the potential to have the least waste of all of the strategies, although in practice that is not often the case. This is because the needed flexibility in both the material and capacity execution of this strategy can be difficult.
The initial thought with a chase strategy is that takt time will have to change all of the time because of the different production rates. However, in practice that is not usually how it is done. Most often, value streams are constructed with a base takt time and operator balance charts to set the staffing level. As the production rate changes, the working time is simply adjusted, by either working more or less hours or adding segments of capacity, (another production line or on half of a line, for instance). The takt time and therefore the pitch stays the same in each configuration.
This slide shows why a chase strategy is difficult in many environments. As the skill level of employees increases, this strategy becomes more and more difficult to execute. The other issues listed here, such as cross training and how to control flow, can be overcome more easily. The staff for peak periods, while a nice fit for a true lean company, is rarely employed by companies who view labor as the largest controllable cost in a company.
This may be a good time to raise the point that the design and manufacturing engineers can, over time, create any bom shape that is desired. Building products from common platforms, use of common components, and finding ways to add customization at the end of the product build process will all help to form a bom shape which drives a more make to order position for the company.
This slide shows that the postponement strategy is really a hybrid strategy, employing both backlog management (in the make to order segment of production) and make to stock management for components.
When we talk here of capacity, we are equating it with the volume from the Executive S &OP plan, not the product mix handled by MPS or Heijunka. The concept, which has been around for a long time, is that it should only be necessary to commit to a production level from suppliers in the long term, thus ensuring an un-interrupted flow of material. Mix should only be committed to at the last moment. This will create some problems if you must deal only at the mix level with your suppliers, because mix is typically more uncertain in the long run than volume is. It may be useful to remember that Toyota has solved these issues by freezing and leveling the production plan for at least two months, and shortening the supply chain so that significant activity can occur in that frozen horizon.
The point here is that you need to select a tool which properly supports the strategy which has been chosen. On long lead time items, where there is no actual demand to drive requirements, none of the options will really work that well, so, again, you should go with the one which is being used for the other items. The higher the mix sensitivity, in other words, the wider the bom is on the bottom, the more the leveling strategy is important, to the point where leveling mix as well as volume becomes the best strategy for managing the material.
Please feel free to contact me to discuss anything that covered in this presentation. Thanks for your interest.