Industrial use big data to improve the technical management of Six Sigma, Six Sigma is the best way that mature supply chain management, Six Sigma universal use in quality control. Six Sigma is a statistical definition of itself, involves a lot of statistical methods and special tools. Observation data from large industrial point of view, in many management program, Six Sigma is the highest priority, "statistical data", which is the industry's best large floor-data of the four scenarios.
Observation data from large industrial point of view, Six Sigma 2 room for improvement:
1, although a Six Sigma used to pay attention to "speak with statistical data", and generally choose a variety of special data analysis tools in Six Sigma, but this is the key statistical data sampling statistics. Statistical analysis focusing on the whole process of sampling that has not reasonable, the Internet can help big data management and use of Six Sigma 150% detection (full review) statistics.
2, Six Sigma lack of public cloud data analysis tools. Six Sigma management methods work in this or expertise, application-specific tools for analysis and statistical quality control must be through training, after confirming a new project, you must do a lot of accurate measurement and data acquisition, data analysis work. Internet data can produce large new projects to improve the standard statistical analysis of patterns, reduce the amount of labor spending on each new project, and then be able to solve a large number of new projects.
Six Sigma as a quality-controlled manner, there are sets of detailed management system, into the definition, measurement, analysis, and manipulation to improve the whole process. The use of cloud computing technology to create a public cloud data analysis platform, a key role in three levels:
1, the quality of comments: commodity display quality information
Sigma CPK able to use standard or value, display information is not the same goods, not the same quality of the whole process, can be used as a new way to solve the selection of projects, looking for less than six sigma standard models of new projects to address law with a new project to solve one by one.
2, look for the harm the quality of the key reasons
Automatic conversion into a Pareto chart, find a good number of the key elements of a total of 90% of the hazardous product quality with a Pareto chart.
3, to improve
Statistical data into elements of the harm caused find favor sake of statistical analysis diagram type, automatic conversion into a highly recommended way of Six Sigma statistical analysis of patterns, such as histograms, statistical analysis, control charts, etc., and from several levels of statistical data analysis carried out.
4, manipulation
Sigma display system software level or value of the information to improve CPK, may be used to carry out long-term plans to improve supervision of the programs and methods.
Scene Six Sigma management application to establish a large information system, the key is to deal with the company the following requirements:
1, how the scientific evidence is not product quality problems: precise measurement of Six Sigma statistical data exceeds the probability of abnormal levels of items to appear only 6.3 parts per million, so that product quality is qualified.
2, the manufacturing industry is an example of what the company quality standards: More than Six Sigma-level product manufacturing industry that is the best specification
3, how to find harm the quality of the key reasons: the quality of harm's sake many, not all of the treatment, the cost of fees must solve problems, to pick more harmful to deal with.
Create a large industrial data system software is based on the company MES and other information management systems to extract manufacturing, testing statistical data, after data cleaning, data stored in the warehouse. Brought together the world capable of fully automatic calculation of statistical indicators using statistical data reception, displaying the results of statistical information analysis chart type way, about the elite team for analysis. Statistical data reception can be statistical data management Kanban (Dashboard), statistics, multidimensional analysis, self-analysis, etc., can also be used to carry out large data mining machine learning algorithms.