American Woodmark is a kitchen and bath cabinet manufacturer based in Winchester, VA. The company is large, with eighteen manufacturing facilities, eighteen builder service centers, and four assembly plants spread across the US. Assembly plants operate on a just-in-time (JIT) production system, which limits their inventory. The management team at the assembly plant in Jackson, GA has identified several opportunities in regards to cabinets being sent to the offline, an extension at the end of each assembly line where failed cabinets are repaired. To measure process quality, the Jackson plant engineers have estimated the percentage of cabinets that pass final inspection on their first try, otherwise known as First Pass Yield (FPY).
According to American Woodmark, they lose, on average, $X per cabinet for every FPY failure. The goal is to identify processes that need to be changed in order to reach the target FPY.
Missing doors and door frames were identified as the most significant contributor to first pass yield problems. A root cause analysis identified several primary reasons why drawer fronts and doors may be missing. They consist of: a part was picked from a wrong inventory slot; the inventory slot was mis-scanned during stocking; the piece is in the facility waiting to be stocked; the part was stocked improperly; the worker at the assembly line didn't have time to attach the part; or the part won't arrive in the facility until later that day. The team decided upon a strategy of prevention and mitigation and thus three solutions are proposed: relabeling inventory slots, shifting the location of door scanners, and changing current stocking procedures.
Relabeling slots: the slot labels in the IC unit do not align with the slots themselves. Since the labels are wider than the slots, as one goes through a bay of shelving, the labels get more and more displaced from what should be the corresponding slot. As such, it is hard to tell which slot label corresponds to which slot, and thus, one could reasonably conclude that these misalignments are the root cause of picking the wrong item. To take care of this issue that is creating missing part opportunity codes, Team 22 decided upon a solution involving the relabeling of all slots.
The relabeling of slots solution will consist of a few changes to the design of the slot label: one, to reduce the font size on the label to make it smaller and two, printing these labels and applying them individually in order to ensure proper alignment with the slot. Currently, the labels spell out the word “Warehouse;” to make them thinner, we suggest using an abbreviation instead, such as “WHS.” Ideally, the labels should be of equal or lesser width than that of the slots. With this new label design and proper alignment, stockers would be able to more consistently scan the barcode on the label of the slot from where they inputted a door. Similarly, pickers would be able to more reliably confirm the slot from where they picked an item.
Changing door scanner location: Come cabinets that hit the offline due to a missing part have unknown short details. This means that there was a missing part for the cabinet, but it got resolved without the expeditor using the Short Recovery system. Because there is no possible way for an expeditor to find a slotted item without going through the Short Recovery system, the only possible source of problem for doors with unknown short details would be not being put on a cabinet in time. Of course, this could be due to several factors, such as the worker’s experience and skill level, but only some causes can be practically addressed.
Currently, workers at the door assembly stage must reach across the cell to scan and confirm the correct door. Sometimes, they do this multiple times before selecting the correct door. This is seen more often with inexperienced workers who are still learning the different door types. Clearly, this can take up some time. While the team can’t change a worker’s experience level, the team can recommend a layout change that would reduce the time it takes to confirm a door. With the current layout, workers must first select a door, walk over to the line, reach over the cabinets on the line, and wave the door around until it gets scanned, at which point the worker can confirm if it is the right door. The extra steps in this process - walking and reaching over the line - could be eliminated by simply moving the scanner to the same side of the line as the door assembly workers.
Stocking solution 1, improved finishing number: Currently, the IC supervisors assign a number of items to be stocked to each individual by dividing it evenly among stockers. The number is calculated by adding the previous day’s leftover items and pallets to the current day items & pallets, and dividing it by the number of stockers they have for that shift. However, this method is not accurately accounting for associated time and capacity constraints for stockers, given that they only work for an 8-hour shift.
The goal in this case is to avoid pallets of doors/drawer fronts/frames sitting on the floor in the receiving area of the facility at the end of the production day. The optimal finishing number for a stocker should ideally be calculated as current day items divided by the number of stockers available. To resolve this issue, Team 22 proposed two solutions. Solution 1 was to hire an additional stocker to the current IC unit stockers if there are a lot of leftover items frequently. If there are only a few leftovers for a few days, then Solution 2 was to set aside extra pay for people that would work after the 8-hour shift to stock more and finish all leftover items.
If the number of pallets/items that comes in a day isn’t completely stocked by the end of the 8-hour shift, then the excess number would be allocated to those who choose to work extra time to finish stocking all the pallets. To motivate them to do so, a pay incentive will be necessary.
Stocking solution 2, improved prioritization of stocking: Currently, American Woodmark prioritizes the pallets to be stocked by using the Orange Cone Method. When expeditors find out that the item needed during that time of production is on a pallet, they put an orange cone on the pallet to indicate to the stockers that they need it. However, this method is not working towards fixing the FPY problem since the cabinet will most definitely hit the offline at least once before they actually get the item. To address this matter, we propose the solution of a better prioritized order of stocking.
Throughout the day shift, pallets of items come in at the receiving dock at different times. For each arrival of pallets, the supervisors in the IC should determine what pallets are needed most by finding the total number of items needed for the same production day from each pallet. Once they determine those numbers from the system, they can place a reusable number sign on top of the pallet with highest priority. Highest priority will start with the number 1 with higher numbers corresponding to lower priorities. Assuming that not many items from different pallets come in needed for the same day of production and that most pallets from previous days have already been stocked, they can make an estimated 5 number signs to simplify the process. If implemented, this would help resolve the majority of cabinets with missing doors or drawer fronts in the offline due to the item being not available for picking but already in the receiving dock. Nevertheless, if for any reason an expeditor still needs an item from a pallet on the dock or coming in later that day, they can keep using the cone method in addition to the number signs. The cost associated with this solution is going to be in terms of labor hours. Team 22 estimated it would take about 30 minutes for the IC supervisor to prioritize the stocking at the beginning of the day. This cost is negligible.
Value and Impact
Combining the three solutions is estimated to result in a 2.50% increase in FPY and yearly savings of $YYY while having an estimated fixed cost of $YY as well as an additional annual cost of $Z. Furthermore, these solutions provide a basis for future improvements that the company can make in order to further improve FPY and the overall efficiency of production.