Deep parsing Industrial Big Data: The Challenge of innovative applications continue

Industry is an integral part of the national economy, the power of a country is strong support behind the competition. China’s industrial highest in the world, but big but not strong. Lack of innovation capability, high-end and high-value products lacking in low-end status in the international division of labor, the Chinese industrial enterprises in urgent need of restructuring and upgrading. We are in the era of big data and digital transformation, data everywhere, the use of data-driven ideas and policies gradually become the consensus in practice. Manufacturing enterprises in the use of big data technology, the production cost can be reduced by 10% -15%, large industrial enterprises of the importance of data is self-evident. Different levels of manufacturing companies in the development process, should take the big data strategies in order from the “Industry 4.0”, “Industrial Internet” and “Made in China 2025” one step closer.
number come from? Industrial Big Data Big Data come from industrial traceability? From all aspects of the product life cycle, including marketing, design, manufacturing, service, and then use the various links, each link will have a big data. “Full” greater life cycle The merged data. Of course, foreign companies, outside the chain “cross-border” Data is the industry big data “can not be ignored,” an important source. The main industrial source big data into three categories: The first category is the production and operation-related business data. Mainly from the traditional enterprise information range is within the enterprise information systems, including traditional industrial design and manufacturing software category, enterprise resource planning (ERP), product lifecycle management (PLM), supply chain management (SCM), customer relationship network storage management (CRM) and environmental management systems (EMS) and so on. Through these companies have accumulated large amounts of information systems product development data, production data, operating data, customer information and data, logistics data and environmental data. The second type is a device object associated data. Refers primarily to industrial equipment and products at certain operating mode things, real-time operation and the covering operation of generating, status conditions, other environmental parameters reflect the collected data equipment and products operating state. Such data is new, the fastest-growing source of large industrial data. It refers to a narrow class data large industrial data, and the presence of large amounts of data, i.e., the time sequence of the industrial equipment and products quickly generated. The third category is the external data. It refers to external Internet source data related to industrial production activities and products business, for example, environmental regulations evaluate corporate environmental performance, market forecasting macro-socialWill economic data. Characteristics (mass, diversity, etc.) differences in industrial Industrial Big Data Big Data Big Data and the Internet in general have great data, the value of having this basis, the timeliness, accuracy, closed loop of four typical characteristics. Industrial Big Data Big Data and Internet industries biggest difference is that big data has a very strong purpose, and more Internet is an association of large data mining, is an analysis of more divergent. Otherwise, both in terms of characteristic data and problems are also different. Unlike large Internet data, core analysis techniques to solve the industry’s big data “3B” problem: 1, Below Surface – occult, that need to discern the meaning behind the industrial environment of big data compared with large Internet data, most the important difference is that the extraction of data characteristics above, the industrial logic of big data focus on the mechanism of the association between physical characteristics and meaning behind the features, and the Internet tend to rely solely on large data mining statistical tools correlation between attributes. 2, Broken – fragmented, the need to avoid intermittent, pay attention to the amount of the aging phase of large Internet data, big data industry pay more attention to the full data, that require the use of application-oriented sample as comprehensive as possible to cover the industrial process changes in the types of conditions to ensure the object can be extracted in order to reflect the true state of comprehensive information from the data. Accordingly, an aspect of industrial large data need to overcome the difficulties caused by fragmentation of data on the rear end of the analysis, feature extraction means using the data into useful information, on the other hand, is the need for the data acquired from the front-end design to the value of demand-oriented development of data standards, and then build a unified data environment in the flow of data and information platform. 3, Bad Quality – poor quality, that is the need to improve data quality, to meet the low fault tolerance data fragmentation defect sources on the other hand also shows that concerns about the quality of the data, that is, the amount of data and can not guarantee the quality of the data, which it could lead to low availability of data, because of the low quality of the data may directly affect the analysis which led to the result can not be used, but the large Internet data is different, it can be done mining, data related only to itself regardless of the significance of the data itself that tapped into what the outcome is what the outcome, the most typical is after the supermarket shopping habits of beer shelf data mining can be placed in the opposite diapers shelves,Regardless of what machine rational logical relationship between them; in other words, compared to the large Internet data is usually not required to push how accurate result, large industrial data for the prediction and analysis of fault-tolerant rate far greater than the Internet big data is much lower. Internet big data during forecasting and decision, just consider the correlation between the two properties is statistically significant, differences between individuals and the noise which can be ignored when the sample size is large enough, thus giving the prediction accuracy of the results will be greatly reduced. For example, when I think 70% of significant Class A movie should be recommended to a user, even if the user does not really like this kind of movie will not cause too serious consequences. However, in an industrial environment, if only through the significant results of the statistical analysis are given, even if only one mistake can have severe consequences. Big Data challenges facing the industry is the first data collection to include data from the network of things and time and space agencies’ information systems attach labels, Quweicunzhen collect as heterologous or even heterogeneous data, but also with historical control data , comprehensive and credible multi-angle test data. The second data storage, to achieve low cost, low power consumption, high reliability goals, to use a redundant configuration, and distribution cloud computing, when storing data classification CONTROL ENGINEERING China Copyright , and adding a tag to facilitate retrieval. Third, data processing, semantic analysis using the context-sensitive, and now on this association context, the international community is a more popular one area. The fourth is the visual presentation , the current computer intelligence have made great progress and development, but can not talk about deep-level mining, the existing data mining algorithms difficult to apply in the industry, what we talk about intelligent road made great progress, but still far away. Considerations large industrial application data in our large data acquisition and processing technology continues to optimize the technology of today, associated with big data processing technology combined with real-time acquisition capabilities of the Internet of Things has demonstrated extraordinary scientific results in more areas of our country, especially in the industrial field of professional industrial data processing techniques for the construction of a large modern factory in our country has brought better technical support, and industrial enterprises want to achieve a better application of industrial transformation of big data must be taken into account as followsFactors: 1, note that a large data plan to improve the degree of well-known large data processing technology combined with numerous types of technology and design level, so companies want to achieve transformation and change using the mightiest industrial data, you need to establish a sound basis for analysis and application environment, the choice of the consumer before the big data industry must consider the basic environmental suitability of enterprises, and a full range of professional industry analysis through large bodies of data to provide a reliable basis for the development to be able to ensure that high-quality technology to get the perfect It implemented. 2, note that analysis of the actual effects of the application of enterprises must carry out a comprehensive analysis of the situation and the actual results may respond to this data after application before applying big data to confirm the brand’s major industrial data acquisition and data processing technology to meet the business actual demand, after application of the industrial enterprise big data can bring real economic benefits and the actual effect is particularly important, so companies can also be a full range of evaluation by large industrial organizations to ensure that data is to lay a good use of technology basis. Must be the basis of various considerations such as the environment and the actual results before the brief application of large industrial data, analyze the benefits and difficulties of industrial large data applications can bring to be able to ensure the use of Big Data industry to achieve perfection, you can also go through professional industrial big data for better counseling agencies to ensure application under this kind of technology can lead to better help the modern enterprise management. Industrial Big Data Big Data Applications development is a process, the ultimate goal is to take advantage of big data, industrial enterprises play a role. Therefore, companies need to keep a cool head, adhere to business applications-driven, in order to maximize the value of data. Enterprise data accumulated in a faster and faster rate increase, many companies will take advantage of the big data technology into the enterprise’s production and management. Big Data applications in industrial enterprises is mainly reflected in three aspects: one is based on the value of data mining products, and related products through secondary data mining , to create new value in the automotive industry , researchers designed a new seat, through analyzing relevant data to identify the owner, in order to ensure the safety of the car. The seat with owner information 360 different types of sensors, can collect and analyze the driver’s weight, pressure, or even multiple messaging like sit on the seat, and they are built in-vehicle system enterMatching rows in order to determine whether the driver is the owner, to decide whether to move their cars. The experimental data show that the recognition accuracy of such saddle up to 98%. Second is to enhance service production is to increase production and improve service type (product) the proportion of services in production values. Mainly in two directions. One is to extend the former, that is, pre-sales stage, through user participation, personalized design fashion Control Engineering Copyright , to attract, guide and lock the user. For example, red collar suit custom clothing, tailored by precise, in other readymade garment shop off the scale of the market, can keep 150% of annual revenue and profit growth , each piece of clothing cost is only 10% higher than the clothing. At the same time after the extension, build customer interaction and product sales by manufacturers to produce sustained value. Apple phone’s hardware configuration is standard, but each apple mobile phone users to install software is personalized, there is the greatest achievement APPStore. Apple Apple by selling the end product is just the beginning, and vendors to establish user connections through APPStore, to meet the needs of individual users, to provide differentiated services, billions of dollars in annual revenue. The third is the business model innovation business model innovation is mainly reflected in two aspects, one is based on big data industry, industrial enterprises outside can provide the kind of innovative business services; the second is in an industrial large data background, you can accept what’s new business services. The best case, by providing innovative business models to gain more customers, to explore more of the blue ocean market, win more profits; at the same time by accepting innovative industrial services, reduce production costs, operational risks. Driven by big data transformation and upgrading of the manufacturing sector, it is to enhance the future of manufacturing production efficiency, improve product quality, save resource consumption, to ensure production safety, the only way to optimize sales services, by industry and the Internet, artificial intelligence, mobile Internet, joint development of technologies such as cloud computing, big data-driven industries of the Internet industry is bound to deeper and deeper into the real economy, has become a new engine of economic era figures.

Related articles