5. Product sales forecast and demand management
Use big data to analyze current demand changes and combinations.
Big data is a good sales analysis tool. Through the multi-dimensional combination of historical data, we can see the proportion and change of regional demand, the market popularity of product categories, the most common combination forms, and the level of consumers. In order to adjust the product strategy and distribution strategy.
In some analysis, we can find that the demand for stationery in cities with more colleges and universities in the beginning of the school season will be much higher, so that we can increase the promotion of dealers in these cities to attract them to order more in the beginning of the school season, and at the same time in the beginning of the school season. The production capacity planning was started one or two months ago to meet the promotion demand.
In terms of product development, product functions and performance are adjusted based on the focus of the consumer group. For example, a few years ago, everyone liked to use music phones, but now everyone is more inclined to use mobile phones to surf the Internet, take pictures and share, etc. The improvement of the camera function of mobile phones is just one thing. Trend, 4G mobile phones also occupy a larger market share. Through big data analysis of some market details, more potential sales opportunities can be found.
6. Production planning and scheduling
The manufacturing industry is faced with a multi-variety and small-batch production model. The refined, automatic, timely and convenient collection of data (MES/DCS) and the variability have led to a dramatic increase in data. In addition, more than ten years of informatization historical data is required for For the fast-responding APS, it is a huge challenge.
The control system of Hangao Tech (SEKO Machinery)'s intelligent stainless steel industrial welded pipe making machinery line can track and record the production data of each welded pipe, such as current size, welding speed, annealing temperature, etc. On this basis, with the introduction of Internet of Things technology, big data can give us more detailed data information, discover the probability of deviation between historical predictions and actual, consider capacity constraints, personnel skill constraints, material availability constraints, tooling and mold constraints, and through intelligent optimization Algorithms, develop pre-planning and scheduling, and monitor the deviation between the plan and the actual on-site, and dynamically adjust the planning and scheduling.
Help us avoid the defects of "portrait" and directly impose group characteristics on individuals (work center data is directly changed to specific data such as equipment, personnel, molds, etc.). Through the correlation analysis of data and monitoring it, we can plan for the future.
Although big data is slightly flawed, as long as it is properly applied, big data will become a powerful weapon for us. Back then, Ford asked what the big data customer needs were? The answer was "a faster horse" instead of the cars that are now popular.
Therefore, in the world of big data, creativity, intuition, adventurous spirit and intellectual ambition are particularly important.
7. Product quality management and analysis
The traditional manufacturing industry is facing the impact of big data. In terms of product research and development, process design, quality management, production and operation, we are eagerly looking forward to the birth of innovative methods to meet the challenges of big data in the industrial context.
For example, in the semiconductor industry, chips undergo many complex processes such as doping, build-up, photolithography, and heat treatment during the production process. Each step must meet extremely demanding physical characteristics. Highly automated equipment is used to process products. At the same time, huge test results were also generated simultaneously.
Is this massive amount of data the burden of the enterprise or the gold mine of the enterprise? If the latter is the case, then how can we quickly find out the key reasons for product yield fluctuations from the "gold mine"? This It is a technical problem that has plagued semiconductor engineers for many years.
After the wafers produced by a semiconductor technology company go through the testing process, a data set containing more than one hundred test items and several million lines of test records is generated every day.
According to the basic requirements of quality management, an indispensable task is to conduct a process capability analysis for more than one hundred test items with different technical specifications.
If we follow the traditional work model, we need to calculate more than one hundred process capability indexes step by step, and evaluate each quality characteristic one by one.
Regardless of the huge and cumbersome workload here, even if someone can solve the problem of calculation, it is difficult to see the correlation between them from the hundreds of process capability indexes, and it is even more difficult to determine the overall quality of the product. There is a comprehensive understanding and summary of performance.
However, if we use the big data quality management analysis platform, in addition to quickly getting a long traditional single indicator process capability analysis report, more importantly, we can also get many new analyses from the same big data set. result.
8. Industrial pollution and environmental protection testing
Based on the Internet of Things, all data in the production process are recorded and monitored, and big data is of great value to environmental protection.
On the Chinese government website, the websites of various ministries and commissions, the official website of PetroChina and Sinopec, the official website of environmental protection organizations, and some special agencies, more and more public welfare and environmental protection data can be inquired, including national air and hydrological data, meteorological data, factory distribution and pollution discharge compliance status Wait for data and so on.
However, these data are too scattered, too professional, lack of analysis, and lack of visualization, and ordinary people cannot understand it. If you can understand and pay attention, big data will become an important means for society to monitor environmental protection.
Baidu's launch of the "National Pollution Monitoring Map" is a good way. Combined with open environmental protection big data, Baidu Maps has added a pollution detection layer. Anyone can use it to view the country and the provinces and cities in their own region, all in environmental protection. The location information, name of the organization, type of emission source, and the latest pollution discharge compliance status announced by the Environmental Protection Agency (including various thermal power plants, state-controlled industrial enterprises, and sewage treatment plants) under the supervision of the Bureau.
You can check the pollution source closest to you, and a reminder will appear, which of the inspection items at the monitoring point exceeds the standard, and how many times it exceeds the standard. This information can be used on real-time social media platforms to inform friends and remind everyone to pay attention to pollution sources and personal safety and health.
The value potential of industrial big data applications is huge. However, there is still a lot of work to be done to realize these values.
One is the issue of the establishment of big data awareness. In the past, there was such big data, but because there was no awareness of big data, and the data analysis methods were insufficient, a lot of real-time data was discarded or shelved, and the potential value of a large amount of data was buried.
Another important issue is the issue of data islands. The data of many industrial enterprises are distributed in various islands in the enterprise, especially in large multinational companies. It is quite difficult to extract these data from the entire enterprise.
Therefore, an important issue for industrial big data applications is integrated applications.
