4. Process design
4.4 Complexity costs drivers
4.4.1 CHANGEOVER PROCESS
Changeover is one of the four elements composing the planned downtime bucket (Table 9). The visible negative effect of the downtime is the stoppage and the production/capacity loss of the line, compared to which the efficiency loss is more insidious. Line efficiency and utilization rate measures are elaborated in section 4.5.2. On top of that, yield loss is the primary cost of changeover. Table 7 summarizes the activities involved in changeover process (Henry, 2012).
Table 7. Activities (3-ups) involved in changeovers
Activity Description
Clean-up The removal of all previous materials, components and product from the line Set-up The adjustment or change out of the line in preparation for the succeeding
product.
Start-up The period of time after the line has restarted but before it is running at normal speed and efficiency
Recipe, format, and label are the three basic changeover types within Heinzβs operational environment.
As can be seen in Figure 18, raw materials are firstly cooked and stored in the kitchen tanks, then after processing (e.g. emulsifying, cooling, etc.) the semi-finished goods are transferred through pipelines and finally filled into bottles. During format and label changes, no clean-up is involved since the recipe and ingredients stay the same. For recipe changes, clean-up usually takes hours emptying the pipe to be ready for the next recipe. Depending on the situation, special cleaning technique or sterilization is required in case of allergic ingredients. In occasional cases, one product range contains variant similar recipes (e.g. Ketchup regular, Ketchup organic and Ketchup less sugar) and recipes changes could be minimized by production sequence planning (e.g. organic ketchup and 50% less sugar ketchup should be produce before regular ketchup, because quality should be assured for organic ketchup not being mixed with regular ketchup but not the other way around).
23
Figure 18. Product flow during the manufacturing process
Changeover drives complexity as it causes loss in terms of both time and costs. However the impact of start-up is often overlooked. Changeover time is defined as the total time spent from the last unit of SKU family A at normal production speed to the first unit of SKU family B reaching the normal speed again. During this period, stoppages, jams, quality rejects occur more frequently, which lowers the overall efficiency. As Figure 19 depicts, the line is stopped during the clean-up and it takes a start-up period to recover to the normal speed interval. If family B is delisted, changeover between A and B could be saved and the line speed will stay at the normal level from tA to tB, instead of dropping to zero and rising back again.
Figure 19. Loss of line speed during and efficiency by changeovers
Normal speed interval
Time Line
speed
Family A Family B Family C Family
D
Start-up Producing Clean-up
Start-up Producing
tA tB
24 4.4.2 YIELD LOSS
Yield loss is the prime driver of changeover costs, which is defined as the amount of scraps and waste that a factory generates while producing products. It is calculated by dividing the material usage variance by the actual spend on raw materials and packaging (Table 8), including any loss is the operation (e.g. the amount of extra product filled over the declared label weight in the final packaged product, spilling on the floor, etc.) (KHC Factory Championship Rule Book, 2016).
Table 8. Yield loss calculation
Actual spend Material Usage Variance Formula Total amount spent
on R&P.
The exceeding value of the material used.
π΄ππ‘π’ππ πππππ π &π β πππππ ππ‘ πππππππ‘πππ π΄ππ‘π’ππ πππππ π &π
4.4.3 WRITE OFF PROCESS
High write off cost is another complexity driver identified. This section describes how the write off is processed.
In the trade contract, a date based on customerβs requirement is agreed, before which the ordered finished goods have to be delivered. This date is called trade Best Before End (BBE), it ensures the customers will have enough time to resell the products within productsβ shelf life. SKUs exceeding their customer trade BBE are no longer available to sell and need to be disposed or destructed, and this activity triggers write off (w/o) cost. Meanwhile, SKUs which have issues meeting trade BBE requirements will be put on the BBE risks list, the risks is posed by the expected number of products to be written off, led by the mismatch of production volume and demand.
Short term
Figure 20. Write off process of S&OP and NOD
25
The write off evaluation and control procedure is described in two interrelated processes: S&OP and NOD (Figure 20). The key responsibility of S&OP in this process is to provide visibility and flag BBE risks to the company, call out attention and action plans to mitigate the risks. If the risky SKUs flagged by S&OP have been selected to be discontinued by SKU rationalization, they will enter the NOD process.
NOD stands for Note of Discontinuation, which is an announcement to the business that a decision has been taken to exit the SKU or SKU family from the portfolio. The NOD process consists of four stages: initiation, indication, execution and finalization. SKU rationalization is one of the reasons to initiate this process. The start of each stage requires the approval of the previous stage and this approval is based on whether the business unit (BU) would like to take the write off costs incurred by this discontinuation.
At this point, S&OP will take the charge again to reduce the forecast write off costs and get the NOD process approved. By doing so, an action plan is to be formulated. Except for monitoring the real time data (e.g. PSI, expiry time, stock leftovers, R&P leftovers, demand forecast, etc.), possible actions could be taken to minimize the write off costs:
Β· Agree base sale beyond trade BBE (need alignment with customer)
Β· Promotional sales
Β· Sell through special (broker) channels
Β· Sampling/Charity
In summary, since write off is one of the complexity drivers in SKU rationalization, the forecast write off value calculated by S&OP is used as an input for candidates selection. While the candidates are decided to be delisted, the NOD takes the control of the overall discontinuation, in which S&OP is still responsible to make sure the write off value is minimized to get the discontinuation finalized.
Figure 21. Mutual relationship of write off and SKU rationalization
Write off differs from other complexity drivers in SKU rationalization because of the mutual relationship shown in Figure 21. In short term, SKU rationalization is a cost driver of write off: if the SKU is decided to be discontinued, all of the corresponding raw and packaging materials, finished goods stock after the exit date have to be disposed. However, in long term, if the write off is caused by the intrinsic complexities of the SKU itself (e.g. demand variation), it is highly possible that the same risks will come over year after year.
SKU rationalization Write off cost
Complexity driver
Cost driver
26 Long term
Although the delisting decision itself incurs write off, SKU rationalization aims at saving costs from complexities. Instead, the situation could be viewed from a long term perspective.
Figure 22. Long term savings of SKU rationalization on write off costs
Most write off costs are caused by reasons such as high demand variability, low forecast accuracy, high MOQ level, sophisticated production techniques, etc. Intrinsic complexities of SKUs are hard to eliminate from the system and will lead to high write off costs year after year if it is not delisted.
As depicted in Figure 22, assume there is an SKU being flagged as high BBE risk at the beginning of 2015, if the rationalization proposal is rejected, the same process will run again in 2016 and BU needs to pay for the write off costs every year following; whereas if the rationalization proposal is approved, the problem ends at the moment the SKU exits the business. Note that no matter if this SKU is to be delisted, actions will be taken to get the forecast value down so that actual write off value is always lower than the forecast. The internal interaction force (Figure 20) already offsets part of the w/o risks.
Hence, in the long term, the savings of SKU rationalization on write off process comes from the avoidance of BBE risks. Given the short-term costs are one-off, the better solution might be to simply eliminate the problem from having to be solved. As for the actual write off value, it is barely predictable as how well the action plan works depends on the situation.