From "Big Data" to "smart data"

Most experts believe that gems and gold can be found in huge amounts of data. Oxford University in the UK has conducted a survey of industry workers around the world, and 2/3 of respondents believe that using data and analytics software can keep them competitive. The question is, how are these "gold diggers" digging gold from such huge data mountains today?

From 3V to 4V

Waiting to discover "gold" refers to the "big data" new technology for recording, storing and analyzing large amounts of data and displaying the results in a suitable form. The most widely discussed topic now is the data that users buy, search for or buy online, or use the global financial and communications network. There are also banks, telecommunications and insurance to increase profits and reduce risk by creating an analytical model of user information and transaction records. The age of large data allows us to explore human behavior and explore the mysteries of human beings, which were largely impossible in the past. We often use tools and terminals to help us gain and experience this feeling.

Because all want to be "Gold Digger", from the large data mining value, now has in-depth analysis, mathematics, statistics, planning skills of the data analyst is hot, there is no enough talent to meet demand. Major U.S. banks and federal agencies are increasingly employing "chief data Officer" (CDO) and data analysts to promote strategic thinking about the collection, analysis, distribution, and application of all functional data across the organization.

Large data has so-called 3V features: "Mass" (Volume), "diversification" (produced) and "Speed" (Velocity). However, the sheer volume of data acquisition is not enough, and the data itself needs to have a higher value, that is, to add a fourth v:value (value) to become 4 v. And after the "Big Data" technology processing (data acquisition, analysis, processing, data display, etc.) will produce higher value.

Intelligent system based on intelligent data

Beer + diapers are data worth digging, while bits and bytes generated from industrial facilities, buildings, energy systems and hospitals are much more valuable and worth digging because they can be used to build smart systems, bits and bytes that are smart data. Let's talk about how smart data sets up an intelligent system.

Terminals are connected and piped, bringing great convenience to people and greatly increasing productivity. But these are not enough, but also need to embody the "intelligent", the realization of intelligent systems. Now we often refer to smartphones, smart meters, smart grids, smart homes, intelligent cities and so on, are all hope that people use equipment and terminals can be automatically programmed according to people's needs, to achieve automation, as far as possible to avoid human intervention.

Such an "intelligent" requires two conditions: the first is "plumbing" (the core of Internet thinking is "piping thinking"), that is, all terminals or nodes are all connected to each other can have "communication" (that is, interaction); the other is that each terminal itself has a "small computer", That is, with a processor chip, you can process and generate "smart data" through software. With these two basic conditions, we can show a certain degree of intelligence.

Take the flush toilet for example. The toilet is a piped toilet, coupled with the second condition, can become a "smart toilet." This can be done in such a way that a microprocessor chip and a biochemical chip (LAB-ON-CHIP,LOC) are installed in the toilet to automatically extract and analyze human excrement, and then the analysis results are piped, such as WiFi, to the doctor, and the Doctor compares the daily analysis data with the data stored beforehand. , give the person who sits on this toilet a reminder of nutrition indicators and physiological indicators, write a prescription if necessary, and remind him to take a drug or go to a hospital for further examination. On the other hand, according to the results of the software analysis of the toilet, you will find out what kind of nutrition specific data, and then through the wireless communication pipeline to the supermarket, the supermarket will be based on the data to choose the right food through the courier service delivery home.

The toilet can also contain various other sensors for "pipe connections", if the use of automatic recording of water consumption, if there is leakage, automatically notify maintenance personnel or property Management department sent to the maintenance, if there is a blockage, will automatically notify the pipeline maintenance personnel to dredge; These will produce a certain amount of data.

We have to understand the amount of these smart data so we can evaluate it correctly; we have to know how the various devices and facilities work, and what sensors and measurement techniques we need to get really important smart data. The decisive factor is not necessarily a large amount of data, but a valuable content.

Such intelligent data can be reflected in various fields. For a large gas turbine, there are hundreds of sensors per second to measure temperature, pressure, flow, gas composition. If people are aware of the physical properties of the facility, and know how to properly analyze the data, it can be very useful advice to power plants to improve the efficiency of electricity use and reduce pollution. The same measures could be applied to wind power, buildings, steel mills and the entire city. In all these areas, data must not only be collected but also understood. The data processed is intelligent, and the results are used to make a business or city smarter.

Algorithms that are appropriate for evaluating these smart data need to be developed. These algorithms can help people to better save energy, better environment, save more cost, and make the equipment run more reliably.

In the future, smart data can help us understand what happens to a smart system all the time, and can tell us why it happens. It can even tell us what happens next and how we should respond. Smart data will change the business model of the enterprise. For example, a multinational company can set up a global repair center, factories in various parts of the world with a large number of sensors and network connection, only in this center to analyze a large number of remote intelligent data, can be remote diagnosis and processing, without the need for technical personnel to the scene. Such a business model is extremely useful for trains, ships, power plants, medical devices, and so on. For example, measured data from the operation of a train can help train drivers to run more smoothly and more energy efficient. Savings can be divided between users and intelligent data providers. This is a winning situation and how to get a good example from the Nuggets in the data mountain.

How large data becomes "smart data"

The data is "big" and doesn't make much sense, and the key is how to best exploit high-value data and use that data as "smart data." There are several ways to do this: Evaluate the value of the data and the value that will be generated, correlate the data with "intellectualization", turn the data into a flexible data structure with contextual implications, and, over time, display a colorful smart data map based on the amount of data collected. In the end, there's no way to think about the difference between big and smart data, since all the data has become intelligent data.

The "plumbing" toilets that the West had invented more than 2000 of years ago opened the internet of things. The advent of internet-based Internet of Things (IoT) heralds the emergence of new innovative devices, new Web Forms, new business models, and a rise in the number of smart data, which will be reflected in a variety of applications. According to today's understanding of the large data concept, is not sufficient, large data must evolve from 3V to 4V, large data must evolve into intelligent data, the entire family and even the entire city is also to "intelligent" stride forward, will have more "nuggets" opportunities.

"Author Zhang Xin, one of the top 500 companies in the world's big High-tech companies as chief scientist"

(Responsible editor: Mengyishan)