In my last two posts, I talked a lot about the free Analysis Platform application, used by Denso and others to dig deeper into big data to understand what exaltiert is happening on your shop floor. I showed you the Digital Native Seven Quality Control Tools (DN7 or DN7QC7). But beyond these seven tools, there are a few more included in the application, which I also would like to show you briefly here.
And, on top of that, I will also briefly show you Seven New Quality Control Tools that are not really part of the application, but aim to dig deeper into problems that are hard to quantify, and many of these tools are more based on relations than on data.
Additional Data Analysis Tools in Analysis Platform
As a reminder, the tool is available for free online on Github, but its support site is only in Japanese, with only a bit of English documentation. But let’s have a look at additional tools beyond the Digital Native Seven Quality Control Tools:
Aggregation Plot (AGP)
Aggregation plots are a combination of bar charts with line diagrams to analyze the results and statistics based on the time period or content of a variable. You could find out, for example, which equipment has the most defective parts. This can be done by year, month, week, day, or hour of the event.
Graphical Lasso (GP)
A graphical lasso (GP) shows you the connections between measurements with adjusted correlations. Like most of these displays, it is interactive and you can zoom in and out and select different points or connections.
Heat Map (HMp)
The heat map (HMp) is a tool to show you data in two different coordinates, usually an x and y position. This shows you trends and differences based on the location. Below is, for example, the heat map of different wavers, with each box representing one chip location.
Stratified Plot (StP)
A stratified plot (StP) shows the histograms of different measurements by machine, part number, line number, batch number, and other variables. This way you can see if certain factors make a difference (e.g., if one machine is more likely to have outliers with your quality).
Scatter Plot (ScP)
You are probably familiar with the scatter plot. This shows you the relation between two measurements. A similar multi-scatter plot (MSP) also shows you all possible combinations of tow variables with different scatter plots.
Principal Component Analysis (PCA)
The principal component analysis (PCA) allows you to monitor the change of many measurements simultaneously, showing you abnormal entries and its causes. This is especially useful for many measurements and can narrow down the causes.
Seven New Quality Control Tools
As mentioned above, there are also Seven New Quality Control Tools listed on the website for the Analysis Platform that look more at relations than at data. It is much more verbal and gives a qualitative understanding of the situation (with the exception of the quantitative matrix data analysis method below). The source for these analyses is mostly brainstorming and expert interviews. It is more for forecasting than for after-the-fact analyses. They are used in Japan, although I am not overly familiar with them myself. Hence, I apologize if the information below is a bit thin. This section here is mostly to inspire, and you will find more if you google the Japanese terms and then use a translation tool. Anyway, these Seven New Quality Control Tools are as follows:
Affinity Projection
The Affinity Projection (親和図法) tries to understand and clarify unknown future problems for complex, often chaotic issues based on affinity and relevance.
Correlation Projection
The Correlation Projection (連関図法) is good for complex relations between factors that can cause problems. It tries to understand the cause and effect of these factors.
Systematic Diagram
The Systematic Diagram (系統図法) tries to arrange factors in a tree shape. It shows which factors affect which other factors to finally affect the outcome.
Matrix Data Analysis Method
The Matrix Data Analysis Method (マトリックスデータ解析法) does a component analysis. This one actually uses data, and tries to compress dimensions of the data into an understandable size. The result is a sort of bull’s-eye graph. It is included in the Data Analysis application Principal Component Analysis (PCA) section (see above), where it seems to be called a “Multi-Scatter Plot.”
Matrix Projection
The Matrix Projection (マトリックス図法) looks at the factors in form of a matrix, listing the degree of relevance. This can give you conclusions on these relationships.
Arrow Diagram Method
The Arrow Diagram Method (アローダイヤグラム法) visualizes the relationships between tasks in the form of arrows, resulting in a network diagram. It is closely related to PERT (Program Evaluation and Review Technique) and helps to find the critical path.
Process Decision Program Chart (PDPC)
The Process Decision Program Chart (PDPC) (PDPC法) helps to understand process decisions and their impact, allowing the prediction of possible obstacles and planning of countermeasures.
Wow. There are a lot of tools shown in this series of three posts. I do hope these will help you to dig through larger data sets to understand what s really going on. Now, go out, improve your quality by analyzing the data, and organize your industry!
PS: Many thanks to Yasuaki Matsunaga from Denso for showing these tools at the APMS Keynote 2025 in Kamakura, Japan!
Discover more from AllAboutLean.com
Subscribe to get the latest posts sent to your email.