Big data as a representative of a new generation of information technology, has begun to make application in industrial design, development, manufacturing, sales, service and other sectors, and to become an important factor to promote the integration of Internet and industrial innovation. The face of the wave of big data, traditional enterprises should take the initiative to grasp the development direction of big data, dig the value of big data, analyze demand preference, improve the production process and improve the internal management level and so on. Industrial enterprises production line at high speed, resulting from the industrial equipment, data acquisition and processing is much larger than data in the enterprise computers and artificially generated, from the data type to see are mostly unstructured data, high-speed operation of the production line is the data real-time requirements are also higher. Therefore, the industrial application of big data problems and challenges faced by no less than the big data applications in the Internet industry, in some cases even more complex.
interactions and transactions between customers accelerate product innovation and industrial enterprises will generate large amounts of data mining and analysis of these dynamic customer data, can help customers to participate in product demand analysis and product design innovation activities for product innovation contribute. Ford set an example in this regard, they will apply big data technology to product innovation and optimization Ford Focus electric car, the car has become a veritable “big data electric cars.” The first generation of the Ford Focus electric car produces large amounts of data while driving and parking. During running CONTROL ENGINEERING China Copyright , the driver continuously updated vehicle acceleration, braking, battery charging and the position information. This is useful for drivers, but also the data back to where Ford engineers to understand the customer’s driving habits, including how, when and where to recharge. Even if the vehicle is at a standstill, it will continue the tire pressure and vehicle data transmitted to the nearest cell system smart phone. This customer-centric big data scenarios have multiple benefits, because the big data to achieve a valuable new product innovation and collaboration. The latest driver to obtain useful information, while engineers in Detroit summary information on driving behavior in order to understand the customer, product improvement program to develop and implement new product innovation. Moreover, power companies and other third-party vendors can also analyze millions of miles of driving data in order to determine the establishment of a new charging stations where, and how to prevent the fragile power grid overload. Product no fault diagnosis and predictionThe absence of the sensor, the introduction of Internet technology makes the product real-time fault diagnosis into reality Control Engineering Copyright , large data applications, modeling and simulation technology is making it possible to predict the dynamics. GE Energy of Atlanta, in the United States monitoring and diagnostics (M & D) center, the data collected over 50 countries worldwide thousands GE gas turbines, 10G day can collect data for the customer, by analyzing sensor signals from the vibration and temperature within the system constant large data stream control Engineering Copyright , these big data analysis for GE’s gas turbine fault diagnosis and provide early warning support. Vestas Wind turbine manufacturers also by the turbine meter data and weather data for cross-analysis, thereby improving the wind turbine layout, thereby increasing the level of power output of the wind turbine and extended service life. Application of large data line modern industrial production line is attached to thousands of small sensors, to detect temperature, pressure, heat, vibration and noise. Because once every few seconds to collect data, use the data analysis can be achieved in many forms, including diagnostic equipment, power consumption analysis, energy analysis, quality accident analysis (including a breach of the provisions of the production, parts failure) and so on. First, improvements in the production process, for use in the production process these large data, you can analyze the entire production process, understand how each link is performed. Once there is a deviation from the standard process flow, it will generate an alarm signal, can more quickly find errors or bottlenecks, but also makes it easier to solve the problem. The use of Big Data technologies Control Engineering Copyright , you can also create a virtual model of industrial production process, simulation and optimization of production processes, when all processes and performance data can be rebuilt in the system, this transparency will help manufacturers improve their production processes. Again, the energy analysis, use of the sensor in the device production process centralized monitoring of all the production processes, it is possible abnormal situations or peak power consumption, thereby can optimize energy consumption in the production process, all processes analysis will greatly reduce energy consumption. Analysis and optimization of large data analysis of the supply chain has many e-commerce companies is an important means to enhance the competitiveness of the supply chain. Advance through big data analysis and forecasting across commoditiesDemand, thereby enhancing the efficiency of distribution and warehousing to ensure that the next day the goods to the customer experience. RFID products such as electronic identification technology, networking technology and mobile Internet technology can help industrial enterprises to obtain a complete big data product supply chain, the use of these data for analysis, will bring substantial warehousing, distribution, greatly enhance the efficiency and cost of sales decline. Haier’s case, Haier supply chain system is perfect, it is a market chain as a link to order information flow as the center, led the movement logistics and cash flow, and integration of global supply chain resources and global user resources. In all aspects of Haier supply chain, customer data, internal data, supplier data are aggregated into the supply chain system, Haier company can continue to supply chain improvement and optimization through large data collection and analysis on the supply chain to ensure that the Haier agile response to customers. Use of sales data, sensor data products and data from the database vendor, industrial manufacturing companies can accurately forecast demand in different regions of the world. Because you can track inventory and sales price, you can buy when prices fall, so the manufacturers can be significant cost savings. If you use data products generated by the sensor, the product know what’s fault, where the need for accessories, they can also predict where and when you need parts. This will greatly reduce inventory, optimize the supply chain. Product sales forecasting and demand management through big data to analyze changes in demand and current combinations. Big Data is an excellent tool for sales analysis, through multi-dimensional combination of historical data, we can see the proportion of regional demand and changes in the degree of popular product categories in the market and the most common form of combination, levels of consumer and other [ 123] CONTROL ENGINEERING China Copyright , in order to adjust the product strategy and Distribution strategy. In some analysis, we can see that more colleges and universities in the city school season demand for stationery would be much higher, so that we can increase sales in these cities dealers in order to attract them to school season, while the school season Two months before the start of a capacity planning to meet the promotional needs. For product development, product functionality, performance tuning by concerns of consumer groups, through the analysis of some of the big data market details, you can find more potential sales opportunities. Production planning and scheduling in manufacturing the face of many varieties of small batch production mode,Fine timely and convenient automatic data collection (MES / DCS) and variability lead to severe increases the amount of data, plus information technology ten years of historical data for the APS requiring fast response times, it is a huge challenge. Large data may give more detailed data we found that historical predictions and the actual deviation probability, given capacity constraints, personnel skills restraint, materials available constraints, tooling constraints by optimizing algorithm intelligence, the development of pre-planning schedule, and monitor planned and actual bias field, dynamic adjustment program scheduling. Help us to avoid the “portrait” of a defect directly to directly impose group identity to the individual (work center data directly change to a specific equipment, personnel, mold and other data). Analysis and monitoring it by correlating the data, we will be able to plan for the future. Product quality management and analysis of the traditional manufacturing industry is facing the impact of big data, in all aspects of product development, process engineering, quality management, production operations and other urgently looking forward to the birth of innovative methods to deal with big data in industrial background challenge. For example in the semiconductor industry in the chip manufacturing process many times will undergo doping, layer by photolithography and heat treatment process complicated process, each step must meet extremely stringent requirements of the physical properties, highly automated equipment in the processed products At the same time, also simultaneously generate a huge test results. Are these huge amounts of data is the company’s burden, or corporate gold mine it? If the latter is the case, then how should restore justice quickly and accurately find the key reasons for fluctuations in product yield from a “gold mine” in it? This is a semiconductor engineers it has been plagued by years of technical problems. Here Leaving aside the huge and cumbersome workload, even if someone can solve computational problems, but it is difficult to process capability index from more than one hundred seen in the correlation between them more difficult overall quality of the product performance has a comprehensive understanding and summary. However, if we use big data quality management analysis platform, in addition to quickly get a long tradition single indicator of process capability analysis reports, more importantly, we can also get a lot of focus on a new analysis of data from the same large result. For the vast amounts of data to be is to refine. For business data, should be done on the basis of real-time analysis, decision support data by moving the end pushed the responsible persons at all levels of decision-making process withData speak, it is no longer a rule of thumb, racking our brains. Big Data to a certain data decision-making power. Data decision-making is based on data for scientific decision-making, and the ability to make data value play. In the era of big data, this capability has become with the previous financial capacity, productive capacity as an indispensable capability.