Part 3: Variations in the Use of Production Schedulers
2013.02.25
Variations in Utilizing a Production Scheduler
To successfully implement a production scheduler, it is important to understand what kinds of utilization methods exist and what characteristics they have.
There are broadly two ways to utilize a production scheduler.
One is to use it as a simulator for delivery date forecasting, material arrangement, personnel planning, etc., and the other is to use it literally as a scheduler for work instructions.
Utilization Method 1: As a Simulator
This utilization method focuses more on visualizing what will happen in the future and taking preemptive measures rather than issuing work instructions.
For example, it can be used in the following ways.
- Predict and respond to delivery dates for tentative orders.
- Predict the expected completion date for confirmed orders. Check the degree of leeway against the delivery date daily, and manage those with decreasing leeway to prevent further delays.
- Check the load on equipment, machinery, and workers, and if there are issues, consider overtime, holiday work, outsourcing arrangements, as well as reviewing personnel placement, promoting multi-skilled workers, and equipment investment.
- Check the schedule of how each material will progress in the future and arrange for timely procurement to ensure timely arrival.
This is similar to the so-called "visualization." However, while normal visualization shows "what is happening now," this case shows "what will happen in the future." This allows for solid measures to be taken while there is still time before problems occur, meaning while there is still freedom to choose various means. Compared to taking measures reactively, it allows for more efficient handling.
Simulation Accuracy
Using a production scheduler enables highly accurate simulations. For example, methods like piling up with infinite capacity in Excel or MRP cannot simulate the differences between high and low production volumes, but with a production scheduler, you can calculate the expected completion date considering the load on equipment and workers with finite capacity. Not only whether it is finite capacity, but also by imposing various operational rules and constraints such as changeovers, worker load, and operation priorities, you can increase accuracy.
If the simulation and actual production are far apart, various problems may arise. For example, if the initially predicted completion date in the simulation does not match the actual date, you may have to deal with it through overtime or outsourcing just before the delivery date. Also, if the actual timing of material consumption is earlier than the simulation, work may not be possible due to material shortages.
Moreover, even if the load on specific machines or workers appears high in the simulation, in actual production, the order of operations may be adjusted to shorten changeovers, and the actual load may not be that high. Conversely, even if the load is low in the simulation, frequent changeovers due to multi-product, small-lot production may occur in reality. If you can accurately simulate considering changeovers, you can avoid unnecessary equipment investment and personnel increases. However, for that, it is necessary to simulate up to the order of operations for each resource.
Even if the accuracy is low, it can be somewhat covered by devising operations, but waste will increase.
For example, in delivery date responses, even if the accuracy is somewhat low, by setting a delivery date with a larger safety margin, the probability of final delivery delay can be reduced. However, this will increase the lead time that can be guaranteed to the customer, which may result in lost opportunities due to not being able to receive orders.
Similarly, in material procurement, it is safe to arrange early with a margin, but this will increase inventory.
However, to achieve high-precision simulation, high-precision data is required. It is also necessary to continuously maintain these data. Additionally, if the production scheduler cannot express the necessary constraints, trying to approximate them can be very challenging.
In summary, when using it as a simulator, even if the accuracy is moderate, you can achieve moderate effects, but if you strive to improve accuracy, the effects will be greater. However, improving accuracy requires corresponding effort and insight. Therefore, it is important to find an appropriate compromise point.
In any case, since you are not faced with a binary choice of "success or failure," it can be said to be a low-risk utilization method.
Utilization Method 2: As a Scheduler
This utilization method, in addition to the use as a simulator mentioned above, involves planning schedules and issuing work instructions to the manufacturing department.
This method can be further divided into two types.
One is the "centralized type," where no discretion is allowed on the shop floor, and manufacturing is done according to instructions. The other is the "decentralized type," where a certain standard is set, and discretion is allowed on the shop floor.
Utilization Method 2-1: Centralized Type
In this method, the production management department plans a feasible schedule and issues work instructions to the manufacturing department based on it. The manufacturing department manufactures according to the work instructions in principle.
The following are the advantages and disadvantages.
◆ Advantages
- The gap between the plan and actual manufacturing becomes smaller, so the gap between the schedule and results also becomes smaller.
- There is a possibility of minimizing production lead time to the extreme.
- For the manufacturing department, unreasonable instructions are not given, so responsibility becomes clear.
- Optimal production throughout the entire factory and even the entire supply chain becomes possible (perhaps).
◆ Disadvantages
- If a plan that the shop floor agrees with cannot be devised, it cannot be operated.
- There is a possibility of resistance and backlash from the manufacturing department.
- Accurate master data and appropriate scheduling logic are required.
- Detailed control in the execution of manufacturing, transportation, and procurement is necessary.
In simple terms, it is a "high-risk, high-return" method, where if successful, tremendous effects can be obtained, but conversely, the risk of failure is also high.
Generally, when introducing a production scheduler, it is common to face the "wall just before full operation," where many problems occur just before full operation, but this method particularly increases that possibility.
To overcome this wall, "accurate package selection," "advanced package utilization technology," and a "solid introduction system" are necessary.
Utilization Method 2-2: Decentralized Type
In this method, the manufacturing department follows the work instructions from the production management department to some extent but has a certain degree of discretion.
For example, "it is okay to change the order of operations within a day."
The advantages and disadvantages are as follows.
◆ Advantages
- The gap between the plan and actual manufacturing becomes (moderately) smaller, so the gap between the schedule and results also becomes (moderately) smaller.
- There is a possibility of (moderately) shortening the production lead time.
- For the manufacturing department, the cases of receiving unreasonable instructions decrease, and a certain degree of freedom is secured, so resistance and backlash are relatively less.
◆ Disadvantages
- Some tuning of the accuracy of master data and scheduling logic is necessary.
- Once satisfied with "moderate," it becomes difficult to aim higher (?).
In other words, it can be said to be a "compromise between ideal and reality" for cases where "completely freeing it would negate the meaning of the schedule, but binding it too tightly is unrealistic."
Which Utilization Method Should Be Chosen?
Based on the characteristics of these utilization methods, determine which method is suitable for your company and make a choice.
When it is said that "introducing a scheduler is difficult," it often refers to the centralized type above. This method is indeed very difficult, and to successfully introduce it, advanced organizational management, appropriate package selection, and utilization technology, accuracy of master data, continuous maintenance, etc., are necessary. If even one of these is lacking, the probability of failure increases. Especially if an inappropriate package is chosen, it will almost certainly fail. The introduction project needs to be carefully advanced, and the period until operation becomes longer. Also, if there are many uncertainties in manufacturing itself and it is often not possible to manufacture as planned, the operation of the centralized type is truly "much labor, little benefit."
However, if it is a workplace that can be completely controlled and has little fluctuation (or can be minimized), there is a possibility of ultimately shortening the lead time.
Even in the case of the decentralized type, a certain amount of effort and labor is necessary. The introduction of a production scheduling system is a project that is difficult to manage because it involves many departments and the interests of stakeholders are intricately intertwined. To successfully introduce it, the same items as the centralized type are necessary. It is not as ultimate, but carelessness is forbidden. Even if discretion is given to the shop floor, if the plan itself is nonsense, there is no point in issuing work instructions. A certain level of accuracy is necessary.
In summary, the safest option is the "simulator," if you work hard, the "decentralized type," and ultimately, the "centralized type."
How to Maximize Introduction Effects
In any utilization method, the production scheduling system is merely a tool. The effect of introducing a production scheduler changes depending on what activities humans perform based on the schedule planned by the production scheduler.
For example, even if it is known that there is a shortage of workers with specific skills, if the information is not made open because it would cause friction, and multi-skilled worker development is not promoted, the expected effects will not be obtained.
Also, even if a splendid plan is devised, it is meaningless if manufacturing and transportation do not follow it. However, in reality, it is quite difficult to faithfully follow detailed and extensive instructions. Some mechanism in the execution system may be necessary.
What Package Should Be Chosen?
As mentioned so far, the higher the accuracy of the schedule, the higher the effect. Therefore, a package that can meticulously express various constraints and limitations of the factory is better. Especially if aiming for the centralized type, considerable compatibility is necessary.
To determine this, it is important to create a prototype and actually evaluate the package. You may think that all Gantt charts look the same, but in reality, each package has its own characteristics and suitability. Choosing simply because "it's cheap," "it's No.1 in achievements," or "the screen is beautiful" is very dangerous.
However, it is quite difficult to know in advance what functions are necessary. It is not uncommon to find out just before full operation when checking the planning results that "this package is no good." Also, various changes occur during operation, and new requirements arise. Rather, that is normal.
Therefore, a package needs "broadness" and "flexibility" to handle situations that were not initially anticipated.
- What changes might occur in this factory in the future?
- What kind of mechanism would be necessary for the system to address this?
- In that case, what would happen with this package?
It is important to use such imagination and carefully select the package.
