Verification of "One-piece Flow Production"
Verification of "One-piece Flow Production"
( Company D's Case )
GD.findi has been developed for use in the planning phase as well as in actual production facilities.
In this example, it will be used to facilitate the product design process.
Introduction – Conveyor belt removal and misinterpretation of a production cell's capabilities
At some point in the history of production innovation, common practice was to remove the conveyor belts running through facilities and to replace them with a small number of manned workstations of W900 x D600 size. This was deemed the era of “cell production”.
Mass production became a thing of the past, LCA (Low Cost Automation) was considered and adopted almost immediately, and each cell was outfitted with simple tools and jigs.
The tendency was for production lots to be produced in their entirety using this natural workflow.Although this “one-piece flow production” may appear to be the ideal form of production, in reality there are a number of obstacles to be overcome which make it at present currently unfeasible. In this example, Company D is an electronics component manufacturing firm that is refining their controller module production facility. In Company D’s case, with the overarching goal of further improving productivity, the effect of each of the changes from conventional conveyor belt-based production to cell production must be carefully considered. However, in spite of the adoption of a cell production system under the guidance of consultants in the field a true understanding of the nature of “one-piece flow production” was not fostered.
With this in mind, by making use of a production simulator, the new design was subjected to a verification process. In this manner, the type of cell would be tested to ensure that it would constitute a viable production method.
The Waste of Over-production
In conventional production strategies, as production plans are set into place, so the issue of wasteful overproduction tends to be shrugged off. However, as cell production focuses around creating the desired lot in “one shot”, the matter of overproduction becomes a significant issue.
In an ideal realization of this process, there would be an invisible conveyor belt running between these cells, providing some sort of communication as well as ensuring that each product is produced one at a time.
In an ideal realization of this process, there would be an invisible conveyor belt running between these cells, providing some sort of communication as well as ensuring that each product is produced one at a time.
Process Segmentation, Cell Specialization, and Insufficient Staff Utilization
Should some worker replace the conveyor belt as a means of moving goods between cells to reduce waste, moving all of the excess at once can constitute an “inter-process buffer”. However such a technique merely serves to increase the size of the inter-process inventory. In addition, upon consideration of the line balance characteristics of each cell it is clear to see that there are variations in the cycle time for each cell, brought about by an assortment of factors. A notable factor in this phenomenon would be a discrepancy between the skills of workers at these stations. As a consequence, a better line balance is achieved by the harmful action of unnaturally distributing processes between these cells of varying quality.
The solution to this problem is to specialize the production cells, reduce the number of staff working at each cell, and perform advanced operations using the skillsets of the experienced workers.
Comparison of Models Using GD.findi
This gives rise to the question as to why “One-piece flow production” is so important, as it is difficult to understand using intuition the inner workings of the process. In theory, the concept is all well and good but in a real production facility space tends to be at a premium, meaning that excess production poses a problem that can be difficult to deal with.
In this case, using a production simulation tool like GD.findi to glean some performance characteristics of each model can help derive an understanding of the phenomena underlying the behavior of methods like one-piece flow production and lot flow production.
Illustrated in the figure below are models and simulation results produced from the “one-piece flow production” and “lot flow production” paradigms that have been discussed up to this point. Although the working time is approximately the same for each cell in each of these models, there is a huge difference in the number of intermediate products generated using each approach.
In GD.findi, in the “floor plan pane” on the left side of the screen, the inventory of each station is represented by a pile of nuggets. In the “production cockpit” window, along with the flow of inventory, the total number of produced products is displayed in addition to the utilization rates of each station. When comparing “one-piece flow production” with “lot flow production”, the difference in the amount of stockpiled nuggets visible in the floor plan panel is quite large. In the “Production Cockpit” window, the stockpiled amount can be seen to steadily decrease. A stable utilization rate for each station can also be observed.
As the lot sizes can change significantly, it the matter of monitoring how the intermediate stockpiled object quantities change becomes trivial. This type of “trial and error” evaluation is clearly unfeasible in an actual production environment.
Verification and Quality Control in Production Simulations and Supply Chain Management
＜Quality Control＞
The factory managers of Company D, upon making the comparison between these two production models, naturally considered the problem of product quality and how changing to one of these models would influence it. The ability to understand the problems within a previous production phase will surely aid in minimizing damage in future iterations. For instance, if a defective part has been produced by some intermediate phase and its existence acknowledged, this knowledge can be used to aid in quality control screening of the final product. In addition, because all of the work-in-progress is handled by a smaller number of workers, it is a simple matter to halt production, resolve any quality problems that may arise, and then resume production at a whim. In “Lot flow production”, because supply generation must be done quickly, the limiting reagent in the production process is in the part delivery phase. A consequence of this is that the lead time in the supply chain is lengthened. ~Supply Chain Management~ In the context of lot flow production, requires that supply generation be done quickly. As the limiting reagent in the production process is in the part delivery phase, a consequence of this is that the lead time in the supply chain is lengthened. Furthermore, if raw material supply is carried out in-house and the size of a customer’s order is small it becomes possible to perform an express delivery of the final product. The link introduced by such a customer enables application of SCM through a virtual conveyor-like construct, and draws attention to members created during the production simulations. ~Unevenness of Production~ “One-piece flow production” as the advantage of creating an instance of the product soon after the start of production. Although the point in time where the final product is produced is somewhat late, this procedure’s forte is that it completes units at a relatively constant rate enabling accurate prediction of lead times. On the other hand, “Lot flow production” is characterized by some unevenness in the work quality on a day-to-day basis. This knowledge is known ahead of time, regardless of the skill of Company D’s workers brought into the equation.
The factory managers of Company D, upon making the comparison between these two production models, naturally considered the problem of product quality and how changing to one of these models would influence it. The ability to understand the problems within a previous production phase will surely aid in minimizing damage in future iterations. For instance, if a defective part has been produced by some intermediate phase and its existence acknowledged, this knowledge can be used to aid in quality control screening of the final product. In addition, because all of the work-in-progress is handled by a smaller number of workers, it is a simple matter to halt production, resolve any quality problems that may arise, and then resume production at a whim. In “Lot flow production”, because supply generation must be done quickly, the limiting reagent in the production process is in the part delivery phase. A consequence of this is that the lead time in the supply chain is lengthened. ~Supply Chain Management~ In the context of lot flow production, requires that supply generation be done quickly. As the limiting reagent in the production process is in the part delivery phase, a consequence of this is that the lead time in the supply chain is lengthened. Furthermore, if raw material supply is carried out in-house and the size of a customer’s order is small it becomes possible to perform an express delivery of the final product. The link introduced by such a customer enables application of SCM through a virtual conveyor-like construct, and draws attention to members created during the production simulations. ~Unevenness of Production~ “One-piece flow production” as the advantage of creating an instance of the product soon after the start of production. Although the point in time where the final product is produced is somewhat late, this procedure’s forte is that it completes units at a relatively constant rate enabling accurate prediction of lead times. On the other hand, “Lot flow production” is characterized by some unevenness in the work quality on a day-to-day basis. This knowledge is known ahead of time, regardless of the skill of Company D’s workers brought into the equation.
The Effect of "One-Piece Flow Production"
In “One-Piece Flow Production”, the product creation phases are visible clearly, and the first product is completed in a relatively short amount of time. This characteristic allows for the designer to see the product and provide input for future iterations of the product without incurring excessive expenditure.
In addition, the same types of inspections can be made by management from a very high-level position regardless of their manufacturing proficiency to pick out “coarse” problems and smooth them out. This will lead to sustainable improvement, as any individual involved in the process will be able to provide input and implement those changes rapidly.
On the other hand, use of the “Lot flow production” paradigm, by performing all of the work at once within each cell the appearance of productivity is increased. This comes at the cost of being unable to view the quality with high granularity, a cost disadvantage, and vulnerable to intervals between lot orders.
The companies who have employed this “one-piece flow production” paradigm are at the forefront of production innovation, as the benefits it brings to the table have allowed them to tap into the minds of their staff to a greater degree.
After Company D started employing the “one-piece flow production” paradigm at their production facilities, productivity has been preserved while reducing the amount of WIP goods. The result is increased flexibility and production power, signifying the advancement of their manufacturing operations.