Toward an Information Society Friendly to Humans and the Earth - Deriving Value from Big Data

[President Endo at NEC iEXPO Kansai 2013]

  This year, we mark the 114th year since NEC’s establishment in 1899. Looking back on it now, I feel that the C&C concept “the integration of Computers and Communications,” which NEC first announced in 1977, remains an important guiding principle even today. This declaration, that the integration of computers and communication can bring about a richer information society, has had a huge impact. This was expanded even further, and in 2007, we established the NEC Group Vision 2017: “To be a leading global company leveraging the power of innovation to realize an information society friendly to humans and the earth.” Now, 36 years after the C&C concept was first proposed, broadband networks and the processing capacity of computers have evolved dramatically. Today, I would like to talk about how we can fully use these capabilities to offer true value.

ICT can provide value in overcoming society’s challenges

  The world’s population has already exceeded 7 billion, and is expected to exceed 9 billion by the year 2050. As the population increases by 30%, the demand for energy is expected to increase by 80%, the demand for food by 70%, and the demand for water by 60%. This is because of continuing urban development. Urban populations will increase from 50% to 70% of the total, bringing about changes in peoples’ lifestyle environments, in which the use of energy, food, and water will increase at an even faster pace than the growth in population. This will no doubt put great pressure on many resources in the future, but these resources simply do not exist in unlimited supply. If, on the other hand, we could get by on 90% of the resources that we are using today, then available resources would increase by 10%. ICT can help us to achieve this goal of 90% resource. In that sense, it would not be an exaggeration to call ICT a “second resource.”

  For the past four or five years, “Cloud computing” has been a hot topic in the world of ICT. The cloud is already entering its second generation; the era of “Big Data”. Big Data refers to the approach of “effectively using ICT platforms comprising computers and networks to generate new value from the data that exists on those networks.”

  One way of using Big Data is to gather, process, and analyze huge volumes of data to make predictions about the future. The new value created through these predictions could demonstrate effects in resolving social issues in a variety of fields, including energy, meteorology, and agriculture.

What is the value to be derived from Big Data?

  From the mid 1990s to the mid 2000s, communication speeds increased by 1,600 times, and the CPU performance of PCs increased by 100 times. The volumes of information distributed on computer networks doubled during that ten-year period. The point is, only one or two percent of that growing volume of data is actually being used. In the future, we need to make full use of that information, and apply the knowledge and experience of users and people in the ICT industry to use that data effectively.

  Big Data will give rise to two types of value: the value derived from huge volumes of data, and the value derived from ICT capabilities. The value derived from huge volumes of data could further be broken down into “discovering implicit information” and “finding patterns and structures within the huge volumes of data accumulated.”

  First, I will give you an example of how gathering large volumes of data enables us to see information that we could not see before. If you look at the windshield wipers on a single car, the only information that you can obtain is whether or not those wipers are operating, and how fast they are moving. If you gather data on the windshield wipers of many cars, however, then you can see how much it is raining in a given region. This is an example of implicit knowledge. Next, I will give you an example of the value of finding patterns and structures in the large volumes of data accumulated. We are now able to capture large volumes of data and effectively extract correlations from the whole. One example of this can be found in programs for Shogi, or Japanese chess. In the past, programmers created algorithms using only partial records of past Shogi games, but now, software has been created using huge volumes of data from past games, and these programs are now able to win against Shogi masters. This means that by using huge volumes of data representing empirical knowledge, we can discover new value.

  There are also two types of value derived from ICT capabilities: real-time, dynamic services, and remote management.

  Let me give you an example of real-time, dynamic services. Imagine that a woman is window shopping, and the camera recognizes that woman’s preferences or favorite color. Product inside the store can then be recommended, in real time, via digital signage. The customer can obtain more information than she normally would just by window shopping, and this creates greater business opportunities for the store. To give you an example of remote management, a doctor in Japan could use a medical robot in Africa to perform an operation halfway around the world. Huge volumes of data that could not be processed in the past, like detailed sensations in a hand, can now be processed remotely, and this could give rise to almost immeasurable value.

NEC’s technologies: Supporting the creation of value

  There are four technologies that support the creation of value: (1) Gathering huge volumes of data; (2) Analyzing the data accumulated; (3) High-speed processing of huge volumes of data; and (4) Network technologies.

  First, when gathering huge volumes of data, we need compact, lightweight, low power consumption sensors to gather the same type of data in large volumes. The vibration sensors developed by NEC are very compact, and have achieved ultra high sensitivity that is about 20x the sensitivity of existing sensors. They can be installed almost anywhere, and can be used in a variety of applications, like monitoring buildings for signs of deterioration. Satellites can also be considered a type of sensor. They can observe conditions in the Earth’s atmosphere with a bird’s-eye view, so this is also an important sensing technology.

  Next is analyzing the data accumulated. NEC is particularly skilled in the fields of face recognition and fingerprint authentication. Both of these technologies are based on finding correlations. Depending on how you look at those correlations, you can find many different types of value. One real-world example of this is the early discovery of abnormalities in plant operations. At a nuclear power plant run by The Chugoku Electric Power Company Inc., we analyzed correlations with patterns from data obtained using 3,500 sensors, and created a model based on the results. The customer then introduced a system that compares real-time data with that modeling data, so that if something appears “out of the ordinary,” then any abnormalities can be discovered quickly, to prevent serious problems before they occur.

  By analyzing correlations, we can begin to see relationships that we could not have even come close to understanding in the past. For example, by using a heterogeneous mixture learning technology, we can find correlations that could not have been quantified before, and that allows us to generate completely new value.

  In the high-speed processing of huge volumes of data, we have the world’s most advanced high-speed stream processing technologies that monitor current conditions and predict situations just before they occur, to enable optimum management. Using these technologies, for example, we can optimize urban transportation or the distribution of energy.

  In the field of network technologies, we are investing efforts into Software-Defined Networking, or SDN, for controlling networks with software. This software controls the flow of data based on changing traffic volumes over networks and available server capacity, in order to dynamically achieve optimum IT and network configurations. For example, we can reduce energy consumption by using specified servers in a data center to their maximum capacity, and then cutting the power to other servers. Immediately after the Great East Japan Earthquake, it was almost impossible to get a call through on a mobile phone. We conducted tests using SDN to change frequency bands, however, and we confirmed that it was possible to dynamically allocate a broad range of frequency bands to mobile phones.

Important security measures

  Once networks can be controlled by software, then infrastructures can be tied into open networks. Up to now, infrastructures in Japan have not been subject to very many attacks, but infrastructures in the United States have been attacked more than 100 times over the past year or two, and cyber terror countermeasures have become a critical aspect of policymaking. We also need to invest more effort into security than ever before.

  As a means of preventing identity fraud, NEC has developed an encryption algorithm called TWINE, which cannot be easily broken, to conceal data obtained through sensors.

  NEC is also working on cyber security countermeasures in collaboration with Interpol, the International Criminal Police Organization, combining NEC’s cyber security solutions with Interpol’s international networks to investigate and analyze threat information.

Conclusion

Our customers have accumulated Big Data in a variety of forms. NEC can use this data to create new value. Our goal is to communicate with customers to find ways of making the best possible use of this data, and to create new value. Thank you very much for your kind attention today.

Top of this page