In recent years, how to improve the quality of automobiles and define the quality indicators of automobiles has become one of the research focuses of various automobile manufacturers. 2018 Year JD power New car quality research (IQS) Display common problems as follows : Inaccurate navigation system, car windows / Fog-prone windshield, built-in Bluetooth phone / Devices often have pairings / Connection problems, audio system is not easy to use, and the built-in voice recognition system is often unrecognizable. The report survey also shows that defects and failures have decreased year by year, but design problems such as vehicle smart interconnection equipment have increased, and common problems such as The improvement in the odor inside the car was the largest, followed by excessive tire noise, no power from the engine after the air conditioner was turned on, excessive wind noise, and excessive fuel consumption.
In short, with the advancement of the technological age, the quality issues and types will change, and the focus of each car factory will also change, but the bottom line is determined by customers or stakeholders. Through different management tools, the company can improve Product performance improves customer satisfaction and increases market share, basically moving toward improving product performance, efficiency, and functionality. The quality management of the automotive industry is based on the ISO/TS standard (as shown in the figure below).
汽車工業的品質管理基於ISO/TS的標準汽車工業的品質管理基於ISO/TS的標準汽車工業的品質管理基於ISO/TS的標準汽車工業的品質管理基於ISO/TS的標準 Quality management in the automotive industry based on ISO/TS standards
There are many links in the automotive industry where data can be generated and analyzed. Taking auto parts as an example, there are accessories such as mirrors, stereos, antennas, etc. Collision parts include sheet metal parts, plastic parts, bumpers, lights, and consumable parts include spark plugs, piston rings, oil pumps, and oil filters. Devices, etc., when each component is inspected, it may be due to different processing procedures, and different processing methods or measurement methods of the operator, which will affect the subsequent assembly quality.
Because the automotive industry must have "the ability to produce consistent product quality", in other words, to reduce the variation in product performance during the production process. Because it may involve many factors in production. We must think about what are the reasons for the variation in production, and what proportion of the variation caused by each reason accounts for the total variation? In order to further analyze the source of variation. This is what we call Source of Variation in quality control, and the main statistical tools it uses are "Analysis Of Variance" and the more in-depth "Variance Component". . Through the operation of JMP software, engineers and technicians do not need to study these relatively complicated mathematical principles in depth, and can quickly find out the variation and complete the analysis report. A more specific case is used below to illustrate.
For example, an auto parts company found that the diameter of the piston rings it produced fluctuated greatly, and the quality control personnel suspected that it was not only related to the production process, but also to the measurement system. Therefore, according to the pre-arranged plan, the measurement results of different measurement personnel measuring different parts at different times during the production process are regularly collected, and then through the statistical analysis of JMP software, the analysis report shown in the figure below is obtained .
ReportReportReportReport report
From the above analysis report, it can be seen quantitatively that the variation of the product itself represented by part accounts for 66.1% of the total variation, and the remaining 33.9% comes from the variation of the measurement system. Among them, the operator represents that the variation of the measuring personnel accounts for 21.3% of the overall variation, operator*part represents the variation caused by the interaction between the measuring personnel and the parts and accounts for 9.3% of the overall variation, and Within represents the other The variation caused by random error accounted for 3.2% of the overall variation. It can also be seen from the variability analysis table that the interaction between measuring personnel and parts is significant. Therefore, if you want to explore the reasons for the variation, you need to look at the operators and their measurement methods and even the hidden reasons behind it, such as whether it is measured by different machines or the impact caused by different time measurements.
But how can we solve this problem? According to the graph of the interaction, it was found that Della The measurement results are obviously different from the measurement trends of other measurement personnel, so it is possible to investigate whether there are other reasons that cause her measurement values to be low, in order to reduce the impact of interaction, In addition, after further on-site investigation by quality control personnel, it was found that the measurement system in the production process not only has different measurement personnel, but also has different measurement equipment. In order to verify our conjecture, we must formulate a more thorough data acquisition plan, and continue to use the statistical analysis of JMP software to obtain the analysis report shown in the figure below.
interaction diagram
Analysis results after adding the device
From the new analysis report, it can be seen that the variation of measurement personnel has been reduced and the variation of measurement equipment accounts for a large proportion. Obviously, we not only need to pay attention to the standardization of skills training for measuring personnel, but also the calibration management of all measuring equipment itself. Only in this way can the variation of piston ring production be effectively and substantially reduced.
Therefore, for the automotive industry, there are still many things worth understanding and exploring the reasons for quality improvement. In the current era of advanced information and abundant data, effective analysis tools (JMP) and production data recording systems are required for timely and effective monitoring and rapid analysis of the real reasons behind it.
