After cloud computing and big data, the Internet of Things has become a new hot topic. The Internet of Things has changed the way we look at the world, changed the way we do business, and even changed our lifestyle. But even the most technologically savvy enterprises admit that it is very difficult to obtain value from the data generated by the Internet of Things, which requires a lot of skills.
Teradata believes that the future of data analysis is "everything can be analyzed", so this conference also released Teradata Listener, a self-help intelligent software with real-time "listening" function. For customers, it can track many sensor and IOT data streams stored around the world and transmit the data to multiple platforms in the analysis ecosystem, so that we can analyze them where the data sources appear.
Aaron Hsin, CEO of Teradata Tian Rui in Greater China
At the same time, Teradata also emphasizes that data islands should be avoided in the construction of data analysis systems. Because a single technology can not meet the needs of comprehensive data analysis, it is necessary to simplify various technical difficulties and create a unified ecological data management system. Simplification is a very important requirement, and any data analysis system should simplify the architecture. Therefore, at this conference, Teradata also updated its Unified Data Architecture (UDA), and introduced the integration of Teradata Data Warehouse, Teradata Aster Analytics and Hadoop system in a single chassis, enabling users to utilize the entire analysis ecosystem management in a smaller data center space.
At this conference, ZDNet interviewed Aaron Hsin, CEO of Teradata Tian Rui in Greater China. The following is an interview record:
The theme of ZDNet:2065 438+05 is to break the big. What does it mean? Does this represent Teradata's conceptual subversion of big data cognition?
Aaron Hsin: The core of the theme of breaking big, I understand, should be "breaking the shackles and restrictions". Both enterprises and individuals should explore and pursue "innovation, differentiation, courage, remarkable progress and outstanding performance."
First, in the era of big data, enterprises must persist in innovation and pursue innovation. No matter how to find a breakthrough in technology, we can actively innovate in business process, business model, organizational structure and enterprise analysis culture. For example, Think Big, which we just acquired last year, has helped us to enhance our consulting, consulting and implementation capabilities for Hadoop, as well as our ability to interact with other analysis platforms. At this conference, we just announced that Think Big has become the first enterprise in the industry to provide comprehensive management services for Hadoop data lake (data resource pool), which will help enterprises to build an ecosystem of data analysis very conveniently and ensure data quality, reliability, real-time performance and daily operation tasks.
Let me emphasize that our Think Big company supports Major Apache? Hadoop？ , including Cloudera, Hortonworks, MapR, Spark, Kafka, NoSQL and other open source technologies, is very comprehensive. More importantly, I'm here to announce for the first time that our Think Big business has been definitely introduced into Greater China, and we are currently completing the staffing.
Second, I think a pragmatic and enterprising culture in data analysis is very important. Among them, the "courage" mentioned in this theme is an important guarantee for enterprises to succeed in big data projects. Many enterprises are hesitant to invest in big data projects, which actually needs more courage to support. Feedback from Teradata and customers shows that it is time for us to take positive action. We also understand that the change of culture may last longer than the change of technology and analysis process, but we have always emphasized that big data started from a young age. I believe you can also see the value of big data soon and see the irreplaceable driving force brought by big data analysis in business change.
ZDNet: At the annual global user conference, Teradata will release new products that attract the attention of the industry. Among the products released this year, which ones do you think have the most highlights?
Aaron Hsin: This year, we made important updates and releases in big data technology, open source technical support and consulting services. I would like to emphasize here that the highlight of this conference should be the ability to analyze the sensor data of the Internet of Things and even realize the analysis of everything. Teradata Listener technology can help customers analyze countless data sources in the Internet of Things, and simplify the difficulty of data analysis by integrating open source technology. Teradata QueryGrid technology can quickly and effectively carry out topic analysis or query diversified big data on a unified data architecture to obtain the information needed by business.
At the same time, the new version of Teradata Aster can directly interact with Hadoop data resource pool or data warehouse platform to help customers explore real-time data, such as analyzing customer paths and consumption patterns in efficient marketing, and so on.
ZDNet: Recently Gartner released 20 16? Ten technology trends that may affect enterprises in 2008, among which everything information technology, Internet of Things and other technologies were selected. When these trends appear in the current development, what do you think of the development trend of technology? If the time is a little longer, according to your observation, which technologies may become more significant technology trends affecting enterprises in the next five years or even 10 years?
Aaron Hsin: We have seen the top ten technology trends, which are strategic megatrends, including the architecture and platform of everything information and the Internet of Things. In fact, I think this is not only a trend, but a new IT reality.
About the informationization of everything, it can be understood that we are in a digital grid, and this environment will produce and use countless information generated by it. In this ocean of data and information, both enterprises and individuals must learn to judge and identify which information can bring strategic value, master how to access these different data sources, and find out the commercial value through various analysis methods and algorithms.
In fact, these predictions are also a portrayal of real IT reality. The most important thing to realize the networking or informatization of everything is sensor technology. In our time, sensor technology combined with large-scale parallel processing ability enables us to measure and analyze almost all phenomena. Advanced instruments enable us to track all changes, such as weather patterns, car driving habits, even the temperature of refrigerators in fast food restaurants and the vital signs of patients in hospitals (or at home). Collect these data into the database, and use a wide range of statistics, analysis and visualization tools to analyze these data in detail.
It is because of these sensors that new data sources appear in our life and work. For example, through RFID readers, we can track and control retail inventory, sample and track medical tests, and prevent fraud. Through GPS position tracker, fleet management, transportation and freight management can be carried out; Through data acquisition sensors, we can collect real-time data for analysis in manufacturing, environmental protection and transportation systems.
For example, Siemens has improved its manufacturing process and product quality by deploying Teradata technology. Siemens realized the data integration from sensors, manufacturing processes, machine-generated data and various source systems for the first time. Dr Michael May, director of business analysis and monitoring in Siemens technology field, said: "Now, we can get the value in the data faster and more effectively. By turning big data into intelligent data, we will be able to optimize product quality and provide better services to our customers. "
I want to say two things about the Internet of Things: According to the 20 14-20 15 annual report on the development of the Internet of Things in China, the collaborative innovation of the Internet of Things technology with emerging information technologies such as cloud computing, big data and mobile Internet has been further deepened, and the "two-way integration" with traditional industries such as agriculture, manufacturing and service industry has been continuously strengthened. The Internet of Things has accelerated its penetration into many fields of economy, society and life, and constantly spawned new changes, new applications and new formats. These are all very gratifying development achievements. With the rapid development of the Internet of Things and the future "Internet of Everything", anyone, anything and anything will be able to connect, which will bring about changes in communication mode, business mode and even development mode.
However, we must emphasize that in order for the Internet of Things to play its role, enterprises must integrate and analyze the sensor data and apply the analysis results to the production process, and the Internet of Things driven by big data is the valuable Internet of Things.
Because the data of Internet of Things is unstructured, the analysis of JSON data is very complicated. In May this year, we announced the first time to realize the native storage of three JSON data formats in the same database, which will provide customers with stronger query performance. By upgrading Teradata database, enterprise users can make full use of the commercial value of JSON data generated by web applications, sensors and Internet of Things machines. Teradata database has the powerful function of analyzing JSON data, operational data and historical business data, and this top-level query performance makes it an analysis hub of the Internet of Things. In addition, the Teradata Listener released at this conference is a self-help intelligent software with real-time "monitoring" function, which can help customers track multiple sensor and IOT data streams stored around the world and transmit the data to multiple platforms in the analysis ecosystem. These are great technological breakthroughs.
In view of the longer-term trend forecast in the future, if we look at IT from a macro perspective, we must first sort out the development of the entire IT industry, and then we can see the future development trend. In the past, since the 1970s and 1980s, the concern for the whole IT industry, whether IT is given by the industry, or the concern of IT suppliers, or the concern of enterprises to set up their own IT departments, is more about "being small and being big". What is a "big T"? The small ones focus on the value that information can embody, and the big ones focus on the application and research and development of technology. This is "the ego is bigger than the t", and more people think that it is only a technical subject, but we should pay attention to that it is not only technology, but also two disciplines, namely information and technology.
With the development of technology, the information value that can be carried by current technology is increasing rapidly. In the future, there will be more opportunities to focus on the information theme and extend the prospect of 10, the next 20 years and 30 years. Especially in the next 30 years, this era will be an era of big I and small T, and more main axes will be on the theme of information. ,
ZDNet: From the experience of Teradata and serving customers, what strategy should you prepare if you ask an enterprise to establish its own big data strategy?
Aaron Hsin: First of all, I suggest that customers should ask themselves a few questions, that is, why should they build their own big data strategy? What business direction needs a data-driven strategy? . Big data strategy should be targeted at specific business scenarios, with clear business scenario objectives, and the ability to control big data should be targeted and mission-oriented.
For example, an enterprise wants to enhance its contribution to customer value and wants to establish a big data strategy, so that it can gain insight from the information of various interaction channels with customers, such as a 360-degree unified customer view, and provide the services or products that this customer needs at the right time, in the right place and in the right way. For another example, by establishing a big data strategy for risk control, financial institutions can discover and judge the risks and hazards faced by their own enterprises, such as guarantee circle analysis. For example, telecom operators can find users who are about to leave the network by establishing a big data strategy optimized for customer service quality, thus improving business support and retaining users.
However, I want to emphasize here that data-driven strategy is not the same as data collection strategy. At present, enterprises should try to avoid "saving but not using". Building big data capabilities is by no means about collecting and storing data.
According to our strategy to help many customers around the world build efficient big data? , I want to share a few keys to success:
First, comprehensive. Enterprises need to identify many different elements that constitute an efficient system from a macro perspective, link different data sets (such as internal and external data streams or information from different functional departments of enterprises), and find out meaningful information through correlation analysis.
Second, take business as the core. The strategic planning of big data should be business-oriented. Big data strategy is not a scientific project, and it must focus on meeting actual business needs.
Third, flexibility. We must consider the future usage, and the big data strategy and big data analysis methodology should avoid common restrictions, such as relying too much on a single technology or a single platform model or over-standardized processes; Since data-driven changes will not spread to the whole enterprise in one step or immediately, it is necessary to realize that value is gradually created and consider the whole evolutionary process when formulating strategies.
Fourth, it is organized and extensible. It is necessary to ensure that the big data strategy can be fully implemented instead of causing another large group of data islands.
Fifth, data analysis and scientific decision. Form an analysis-oriented way of thinking and cultivate a real data-driven culture.
That's what Bian Xiao shared for you about the future of big data analysis. Everything can be analyzed. For more information, you can pay attention to Global Ivy and share more dry goods.