Toward an Information Society Friendly to Humans and the Earth – Managing information to empower the society –

Today I would like to speak on the theme of "Toward an Information Society Friendly to Humans and the Earth - Managing information to empower the society". When we talk of being "friendly to humans" we look to the power of ICT to enrich people's lives by providing safe, secure and convenient services to anyone, anytime and anywhere. The power of ICT also allows us to be "friendly to the earth" through the efficient use of limited resources and co-existence with the global environment to bring about sustainable development. We believe that ICT capabilities are vital if we are to live enriched lives in harmony with the world around us, and it is the use of cloud computing and big data in particular that will become increasingly more important.

Last year, the global population topped 7 billion people, and this figure is expected to grow to more than 9 billion by 2050. Around half of the world presently lives in urban areas, but it is thought that this figure could increase to 70% by 2050. Besides the obvious effects of urbanization, a 1.3-fold increase in the population by 2050 is predicted to bring a 1.8-fold increase in the demand for energy, a 1.6-fold increase in the demand for water, a 1.7-fold increase in the demand for food and a 1.5 fold increase in greenhouse gas emissions. It is imperative that we put in place smart social infrastructure based on ICT that maximizes our limited resources.

The rate at which we use resources can be worked out using the calculation: resources over usage volume multiplied by the arbitrary figureα. Given that the world's population is certain to increase, we have little hope of building a safe and secure future for 9 billion people unless we either uncover new resources, the numerator in our calculation, or reduce the value of our denominator. The real issue, however, is that it is virtually impossible to reduce resource usage while our population continues to grow. Therefore, the only way we can bring down the value of our denominator is to reduce the arbitrary figure, α, to below 1, and it is here that we find the perfect role for ICT. It is an important element in efforts to solve the issues we face. ICT can be used to optimize energy usage and initiate energy savings, or be deployed in supply chain management to increase efficiency in food utilization.

The key is to use big data to collectively process hereto unimaginable volumes of data to uncover inherent value. As a result of 20 years of rapid development in IT and network platforms, we are now able to make use of big data to create new value.

Information consists of "explicit" and "implicit" information. When a problem occurs we tend to use readily available or explicit information to pinpoint the cause of the problem. However, when background or implicit information is actually collected and analysed "fundamental issues" often come into view to challenge our original assumptions. This is a key benefit of using big data.

Allow me to use a car as an example of the value of implicit information. Cars are fitted with a large number of sensors, but here let's just consider the windshield wiper sensors. Information obtained directly from these sensors include whether the wiper is operating and at what speed. This is explicit information. Information gathered from wipers on a single vehicle can only be explicit information, but if we collected wiper information from multiple vehicles across a wide area, other information would emerge, such as where rain is falling, which way rain clouds are moving and the changes in rainfall over time. This is what is known as implicit information. In this way, data consists of clear and hidden elements, and by gathering large volumes of data and analysing it for hidden information we can create new and different value.

Creating value from big data requires that we "gather large volumes of data and information", then "analyze this data and information" to "create value from it". Allow me to explain this process step by step.

First, we "gather large volumes of data and information".
Data is gathered from sensors, cameras, smart devices, and even satellites, but effective data collection requires that each of these have some important elements. Sensors should be compact, lightweight and energy saving, while cameras require technology that extracts the desired information from stored image data. Other important elements are user interface design methods and formatting when converting electronic data. Security measures when gathering data are also critical. Cyber attacks that replace data during the data gathering process could adversely affect any social infrastructure being controlled by this data.

At NEC we are actively pursuing the development of these technologies. In the sensor field we have developed a compact and lightweight ultrasensitive vibration sensor. This revolutionary product is just 8.5 millimeters square. It is 20-times more sensitive than conventional sensors and one-tenth the cost. Sensing minute vibrations, it can detect water leakage on pipelines, or even the sound of a pulse when fitted to the wrist. It is also effective in preventing theft or detecting degradation in buildings, roads and infrastructure.

In the area of security technology, NEC has developed, a new encryption algorithm that enables high-speed processing in all environments, including small, dedicated encryption devices and software processing on microcomputer servers. This technology can also be fitted to compact sensors operating on the world's smallest calculation resources as an effective measure against data tampering and theft.

The next step in the process of creating new value is "analysing data and information". NEC provides a variety of analysis engines. Invariant analysis automatically detects correlations in data. Heterogeneous mixture learning automatically detects completely different patterns and rules in large volumes of data. Facial image analysis automatically detects faces within images. Behavior analysis analyzes the position, time and movement history of people and objects, while recognizing textual entailment determines whether two texts contain the same meaning.

Invariant analysis technology analyzes the correlations within large volumes of data to detect rules and regularities. Take for example a large-scale factory, where sensors collect data by monitoring such items as flow of goods and temperature levels. Rules and patterns in this data are accumulated as "experiences" to create modeling data. This modeling data is compared with current conditions in real time to enable early detection of irregularities or changes in the ordinary operation of the factory to prevent trouble before it happens.

Mixed type learning technology collects and analyzes not only surface data but also background data. Say we want to gauge energy consumption in office buildings. This technology collects and analyzes large volumes of data on seemingly unrelated topics, such as temperature, time, humidity, PC operating rates, ratio of men to women, light levels and total employee numbers. Patterns are then detected from this data to enable us to accurately predict future energy consumption levels to within 2-3%.

The face image analysis engine features a database for defining facial data obtained through the collection of a large volume of facial images. The technology utilizes this database to detect facial data from crowd surveillance images to create high-resolution images that display such details as eyes and the shape of the nose.

The last step in the process is to "create new value" from the data analysed with these technologies. At NEC we are delivering "analysis cloud" and "data analysis and verification" services. These services filter analysis derived from the Big Data platform and analysis engines, and convert the results into valuable information for our customers. Making full use of big data enables us to improve the accuracy of traffic forecasts, predict or detect disasters and optimize production lines. We can also create new value in a diverse range of other fields, including distribution, environmental protection, medicine and energy.

Finally, I would like to talk about the development of real time, dynamic ICT platforms. A keyword in this development is "Software Defined Networking (SDN)". This technology enables real-time service delivery without the collection and processing of data by a data center. Known as "streamline computing", SDN is a new computing concept that solves dynamic issues with an intelligent network that gauges the size and destination of streamed data in real time to determine optimum routes. As an indispensible solution for SDN, NEC has developed and is selling "ProgrammableFlow ", the world's first OpenFlow compatible product. The solution has attracted attention and praise around the world, beating out considerable competition to win the Grand Prize at network products expo, Interop 2012.

ProgrammableFlow has already been deployed in customer data centers with remarkable success. Demonstration testing on closed networks has verified an 80% reduction in power consumption and a 98% reduction in switchover time in the event of failure.

ProgrammableFlow is currently being used in demonstration testing being conducted with the Japanese Ministry of Internal Affairs and Communications to dynamically change network configuration and prioritize emergency communications, such as voice calls, email and data recording, in times of disaster. This service will also owe much to SDN technology and its ability to control networks using software without the need for hardware modifications. As the volume of data handled over networks continues to grow, SDN will undoubtedly become a key technology of the future.

Today I have talked about ways to create value through the collection and analysis of big data. The creation of value from big data is imperative if we are to build an information society friendly to humans and the earth, and we believe that "real time" and "dynamic" will be keywords in our efforts to achieve this goal. We are about to embark on a new era; an era in which large volumes of data will be collected and processed in real time to dynamically deliver outstanding solutions to our customers.

We always welcome honest comments and suggestions on NEC solutions and are committed to solving any issues that may arise. Please don't hesitate to provide us with feedback. Thank you for listening today.