Six sigmaThe problem-solving procedure page mainly explains the problem-solving QC story, but this page does not mainly explain the problem-solving QC story. Problem-solving QC stories can handle as many problems as "all problems". On the other hand, task-achieving QC stories are not so versatile. The cause seems to be that there are multiple types of "tasks" rather than the deficiencies in the task-achieving QC story.
On the Problem Solving and task achievement page, we talk about how to use "problems" and "problems" that are generally confused. It seems that the issues can be further divided into three.
Depending on the type of task, the appropriate procedure will also differ.
It is a case where the task can be achieved by fixing something bad (problem) that has existed in the past.
If you notice this kind of problem, it is easier to proceed with a problem-solving QC story , thinking that it is problem-solving rather than problem-solving .
Although there are words that express goals such as "doubling sales" and "development of new fields," these are vague issues. This may be the most difficult case. You need to look at the current situation and causality to see what you can do to achieve the task.
Six Sigma 's DMADV is a good challenge.
DMADVs are more likely to analyze (think) and then try to do it. There are 5 stages: Define, Measurement, Analysis, Design, and Verify. At the 4th stage, "Design", we will decide to do it concretely.
This is an issue that is premised on creating something, such as "building a building", "introducing an accounting processing system", and "developing a new product". In the field of artificial intelligence (AI) , "introduction of visual inspection AI", "introduction of anomaly detection AI", etc. are applicable.
For this task, we have decided what to make, but not the details. In the case of a building, we will proceed while deciding what kind of place to stand and what kind of building to make. Accurately grasping customer requirements and making on time are the most important issues. Compared to other challenges, the "design" stage is especially important.
Not limited to PMBoK, which is the standard for project management, in what is called "project management", the order is start-up, planning, execution, monitoring / control , and so on. The order of project management is suitable for the task of making something.
By the way, it is a task that you can proceed after considering countermeasures in the problem-solving procedure and you can freely decide the policy, and the procedure after deciding the policy is the same as the task of making something.
The task of making something, even if there are other options, can go on without thinking about it.
For example, you may be advancing to "build a house" without considering "is it really the best to build a house?"
It's tempting to think that it's just a matter of thinking and choosing, but in the real world, it's often not possible or extremely difficult to choose.
I wrote above that there are three types of tasks, and the appropriate way to proceed is different, but the task achievement type QC story is not included in this.
In my opinion, the conclusion is that the task-achieving QC story can be used for any of the three types. However, for all, it is harder to use than other methodologies.
First, the task-achieving QC story is as follows.
"Selection of theme", "confirmation of effect", and "establishment of standardization and management" are the same as the problem-solving type.
It is hard to think of a distinction between grasping the current situation and factor analysis . Therefore, it is difficult to analyze the structure of the event and then formulate a policy based on it, and it is rather easy to come up with a policy such as an idea game.
For tasks where the strategy is decided from the beginning, the task-achieving QC story fits fairly well.
However, because it fits well, it becomes impossible to "eliminate the initial assumptions and tackle the theme comprehensively and systematically" like the problem-solving QC story. It is practically similar to the policy execution type QC story .
The problem-solving procedure is basically an activity that tries to fix the bad parts of the existing one.
On the other hand, as a measure to achieve the task, changing the whole existing one to another mechanism may be an option.
It is a measure that makes it possible to perform the procedure that was done on paper only on the computer.
The advantage is that it is an order of magnitude more effective than working on some of the existing ones, leading to more of a "reform" than an "improvement".
The disadvantage is that the effort and cost to realize it can easily be orders of magnitude. In addition, there is a possibility that problems unique to the new mechanism will occur.
Reengineering was advocated in the 1990s. It means recreating the way of doing business according to the situation.
Continuing a difficult job can be a daunting task, but it can be forced to decline as the times change. If a company wants to develop sustainably, it is also necessary to dismantle and recreate the shape once it was created.
The best feature of re-engineering is "re (again)". Rather than so-called "improvement" and "new launch", reengineering aims to change operations from a more fundamental point.
For re-engineering, using the concept of value engineering and QFD (Quality Function Development) , "What is this to achieve in the first place?", "What kind of method would you like to do the same thing differently?" Do you have any? "
In addition, the explanation of re-engineering may include the same improvement activity as "fixing the bad points of the existing one" or starting a new business.
Problem-solving QC story in, machine learning may approach is useful. However, unlike when used as artificial intelligence (AI), it is not often used as "reasoning" or "prediction". If anything, it is used to understand the whole picture of data and phenomena by looking at the relationship between the model and data during training.
On the other hand, if the goal is to achieve the task, you may want to use inference or prediction techniques as a measure, depending on the content of the task. This story is also in data science for problem solving and task achievement .
NEXT Six sigma
