Data is the key to the future. But in order to unlock that future, you need to know how to use that key. The promise of artificial intelligence (AI) and analytics is that these tools will allow us to exploit the wealth of data that is available today and the even greater amounts of data that will be available in the future. At NEC, we have been a leading force driving the digital transformation, devising new AI technologies to help users take advantage of the explosion in data. By incorporating that technology in new and innovative products and services, as well as in human resources, and by zeroing in on specific social issues, we are able to offer resources that will help our customer solve those issues. AI technology is expected to advance rapidly in the years to come with a focus on the three basic frameworks that form the foundation for AI solutions — visualization, analysis, and prescription. At the same time, more efforts will be made to make AI more accountable, especially in terms of processing transparency and social acceptability. Finally, rapid development of small data learning technology will enable companies to more easily exploit deep learning technology and quickly get smart, practical fast-learning solutions up and running.

While Artificial Intelligence (AI) can enrich our lives, it may also lead to human rights issues such as the invasion of privacy and/or discrimination depending on how it is utilized. Anticipating and dealing with these issues is now a source of lively debate in government, academia, and business. At NEC, we have focused particular attention on these issues and in April 2019 finalized a set of principles — “NEC Group AI and Human Rights Principles” — which we introduce in this paper, outlining both the background and our commitment to putting these principles into practice.

As we move towards the “Super Smart Society” outlined in the Society 5.0 initiative proposed by the Japanese government in the 5th Science and Technology Basic Plan, new human resources are expected to play an active role in solving various social issues by utilizing artificial intelligence (AI) and thereby creating new value. Today, a worldwide shortage of AI specialists has pushed companies into intense competition with each other to secure competent AI specialists. To deal with this problem, a more aggressive approach to developing human resources specializing in AI is required. At NEC Group, we have been working hard to develop human resources specializing in AI since 2013 and many future AI specialists have passed through our workplaces. Based on a case study of NEC Academy for AI, this paper shows how we develop our human resources in the age of AI.

To ensure the success of AI operation, it is essential to establish a collaboration between various specialists from data scientists to applications developers. NEC Advanced Analytics Platform (AAPF) is an AI operation platform that supports such collaborations. AAPF allows the use of analytics tools employed worldwide as open-source software (OSS) as well as NEC the WISE technologies, including Heterogeneous Mixture Learning. The container technology facilitates the creation of an analytics environment that meets the diverse needs of individual users. From data analytics environments to, AI development environments, AI execution platforms, and learning environments aimed at human resource cultivation, AAPF is currently being used in a wide range of applications and is supporting the implementation of AI in the field of business. This paper introduces AAPF and examples of its applications.

One of the key technologies of “NEC the WISE” , the leading AI technologies of NEC Corporation, is a group of technologies known as the “Fingerprint of Things” recognition technology that can identify individual objects based on the fine patterns that appear on the surfaces of industrial products and parts during their production processes. It thereby enables the individual identification of various objects that do not access traditional processing such as the attaching of ID tags or laser markings. NEC started the provision of the individual ID function based on the Fingerprint of Things technology in October 2016 under the name of GAZIRU Individual Identification Service, which is currently available via on-premise implementation in the GAZIRU Individual Identification Engine. The present paper introduces actual cases that use the individual ID function.

The use of artificial intelligence (AI) in product inspection applications is becoming increasingly common. This paper examines a joint effort between NEC and Nippon Electro-Sensory Devices (NED) — a manufacturer specializing in line sensor systems — to create a defective product detection system that incorporates NEC’s RAPID machine learning in image processing software used in image inspection systems. In addition to providing customer value which makes it easy to build an image inspection system incorporating machine learning, we have incorporated other cutting-edge NEC technology into NED’s products with a view to deploying these solutions in new fields and promoting digitalization in manufacturing.

Expectations for future driving technologies have recently been growing in the mobility domain; one of these is in the practical use of drive recorders for remote surveillance and another is that concerning remote driving systems. However, as vehicles keep on moving, the communication bandwidths of mobile networks and the amount of transmitted camera video data may change in a complex manner. This issue causes the videos to be disturbed when the real-time surveillance of videos from multiple vehicle-mounted cameras are processed. In order to solve this issue, NEC Corporation has developed the communication prediction/control technology that integrates the communication prediction and the communication control technologies. Communication prediction technology predicts changes in the communication bandwidths, and communication control technology detects important communications from multiple cameras and automatically optimizes the communication bandwidths using a rule-based AI. The present paper introduces a remote surveillance solution that enables real-time surveillance of videos from multiple vehicle-mounted cameras using the developed technology.

Over the past few years, both the Japanese government and Japanese corporations have been pushing for work style reform with a focus on flexibility and diversity, as well as health management. NEC’s Emotion Analysis Solution reads, quantifies and visualizes a person’s emotional state. This data can then be used to develop an appropriate policy, enabling organizations to implement operating and workplace environments optimized for each employee. In this paper, we describe the system configuration of NEC’s Emotion Analysis Solution and application examples, as well as future prospects.

Office security is widely regarded as one of the most important issues facing businesses today. While traditional lock and key security has not been abandoned, a wide range of authentication methods are now being used in the modern workplace, including key fobs, ID cards, and passwords. Unfortunately, these enhanced security systems tend to inconvenience users, and can often be confusing and complicated, as well as leading to more complex management and administration. Further complicating the situation is the threat posed by theft and use of other people’s ID cards and passwords. NEC’s Facial Recognition Solution for Offices offers a comprehensive suite of facial recognition products and services that allows companies to achieve a secure and convenient office environment by standardizing intra-office authentication with facial recognition.

The auto response solution, with which the AI responds to inquiries, is recently attracting attention as one of the applications of AI technology that can make up for human resource shortages and support diversified ways of working. It is especially notable that the use of chats as a means of inquiry other than the telephone and e-mail is increasing because of the ease of use, and the chatbots are expanding their usage scenarios as a familiar AI technology. This paper introduces an outline of an auto response solution that can give highly accurate answers to inquiries by utilizing the Recognizing Textual Entailment technology. This procedure can recognize the diverse expressions of natural text.

In today’s digital world where globalization and rapidly evolving technology have created a tumultuous and unpredictable business environment, many enterprises are struggling to keep their footing in the face of upstart competitors, shifting industry boundaries, and shrinking talent pools. Under such conditions, workforce management can no longer be taken for granted. Hiring, firing, scheduling, performance assessments, and day-to-day operations need to be optimized to ensure that enterprises function at their best and that their employees are happy and productive. In Japan, a tightening labor market makes the situation particularly acute. Finding and keeping the right talent for the right positions is a growing challenge and companies must compete with one another to attract the best employees. This makes it imperative that businesses provide environments and systems that will allow staff to concentrate on creating new value and generating new business opportunities. At NEC, we have developed a new solution that exploits the astonishing advances in AI technology to take over many of the daily routine tasks and labor-intensive operations. In consequence, white-collar workers are called upon to perform, leaving them free to focus on decision-making and new value creation. In this paper, we will examine this technology in detail and highlight several test cases that demonstrate its effectiveness.

NEC has since 1951 been accumulating an abundant business experience with energy utility companies. Based on this experience, we have developed various energy management projects aimed at achieving an efficient and sustainable society. As one of these projects, we have developed and started providing an “NEC Cloud Service for Energy Resource Aggregation” (hereinafter referred to as the RA Cloud Service,) which is an energy management service employing AI technology. In 2016, the Ministry of Economy, Trade and Industry publicly offered the “Virtual Power Plant (VPP) Construction Demonstration Project”. The RA Cloud Service is targeting companies that participate in this project in order to make VPP a practical business solution of the future. RA Cloud Service supports the customers in enabling the best use of AI technologies to control and optimize the energy facilities owned by the consumer side in order to meet the Demand Response (DR) scheme.

The growth in hospitalization periods is raising concern with regard to increased social security expenses. Delays in hospital discharge of patients are mainly caused by two issues. These are the extra treatment required due to changes in the condition of inpatients and also by the delay of hospital discharge scheduling due to inefficient office procedures. This paper introduces efforts being made regarding agitation and aspiration pneumonia issues. These are the main factors causing delays due to the need for extra treatments. NEC Corporation has developed an AI technology that detects agitation and aspiration pneumonia based on the learning of electronic medical records and vital signs. The use of AI technology in this topic, which has been considered as being difficult to detect before it actually occurs, enables advance intervention by medical personnel and also supports the prospect of earlier patient discharge.

The low birthrate and the increase in population ageing in Japan continues to accelerate into the future and the national medical costs are expected to grow to 66.7 trillion yen by 2040. As the increase in medical expenses and the decrease in the working generations are becoming social issues, an awareness of disease prevention and healthcare management are becoming more and more important. This paper introduces the effective, scientific approach of NEC Corporation in challenging the issue of prevention/healthcare management. It is based on the analyses of previously collected data using AI, instead of blindly adopting random measures. The procedure described below attempts to identify the causal relationship between lifestyles and laboratory values by combining the traditional statistical techniques and AI (heterogeneous mixture learning).

In a world where the boundary between the real and digital worlds is becoming increasingly fuzzy, effective marketing requires the development of richer and more powerful “consumer insight” that can be used to motivate and trigger consumer behavior. At NEC, we are pursuing transdisciplinary solutions to achieve consumer insight that draws from deeper layers of data which is essential for implementing today’s market strategies; a more comprehensive understanding of consumer moods, feelings, desires, and aspirations. This study focuses on a joint effort by NEC and Macromill to develop a sophisticated transdisciplinary suite of tools for gleaning consumer insights by combining NEC’s AI technology with the wide range of consumer data possessed by Macromill.

The result of a unique collaboration between NEC’s AI technology and a renowned bean-to-bar craft chocolate maker, “Anokorowa CHOCOLATE” captures the flavor of some of the most exceptional years in the past half century. Our AI analyzed a massive number of words in newspaper articles and converted results into multiple taste index values, which were then used to create five different chocolates with distinctive flavors that capture the mood of the time. This paper introduces the methods used to convert words into taste index values and discusses the potential for product co-creation between humans and AI.

As the recent reduction in the working population has been tending to lead to shortages of labor in the retail industry, labor saving and unattended sales procedures using AI are strongly anticipated, particularly in the payment operations (register checkout) with high operational loads. This paper describes the heterogeneous object recognition technology that supports unattended payments based on the simultaneous image recognition of objects. These may vary from industrial products such as packaged retail products to natural goods such as daily-delivered products and perishable products. We also introduce an image recognition-based POS system that makes use of this technology, allowing customers to perform fast payments by simply placing multiple products even when arranged randomly.

Optical fibers used in the optical communication technology industry to support worldwide high-speed Internet have an information collection function that captures changes in the surrounding environments as well as performing the basic function of information transmission. NEC Corporation is currently conducting R&D of optical fiber sensing technologies that utilize optical communication infrastructures as sensors for visualizing the real world. It is doing this by combining the world’s top level optical communication technology cultivated via its experience in the submarine optical cable systems together with the latest AI technologies. The present paper introduces basic principles of the optical fiber sensor with system configuration and discusses three case applications: intrusion detection of facilities, surveillance of road traffic flow and bridge deterioration detection.

Artificial intelligence (AI) is now being used to automate a wide range of tasks. Typically, automation is achieved by setting optimization indices for the levels of acceptability or unacceptability in target tasks and having the AI automatically search for the decision that maximizes or minimizes those indices (optimal solutions in mathematical optimization) as required. However, with tasks that require a certain level of skill and expertise — what we might call “expert-dependent” tasks — setting optimization indices is more difficult, making these tasks harder to automate. This paper introduces NEC’s intention learning technology which learns intentions from the decision-making history data of various experts. These intentions can be used as optimization indices, enabling the AI to imitate the decision-making processes of experts in the task being automated and making it possible to automate tasks that would normally require a human expert to accomplish.

We can represent numerous analytics targets as graphs, where a “graph” consists of nodes representing data points and edges representing their various relationships. Diagnosing a patient, for example, not only depends on the patient’s vitals and demographic information but also on the same information about their relatives, the information about the hospitals they have visited. Predicting the future performance of a start-up not only depends on its business metrics, but also the relationships with other companies and individuals, their networks and expertise. By pushing this insight further, a team comprising NLE (NEC Laboratories Europe GmbH) and CRL (Central Research Laboratories, NEC Corporation) are in pursuit of Graph-based Relational Learning, which frames the world into graphs, and achieves groundbreaking new technologies for applications in various business domains. Does it not only improve the performance of node classification, but also delivers link prediction and graph classification tasks, while integrating multi-modal and incomplete data sources, and further providing explainability to complex AI models.

Using deep learning to extract and compare statistics and other char acteristics of time-series data, retrieval-based time-series data analysis technology can assess and pass judgment on a given set of conditions with high speed and high precision. Unlike conventional data analysis techniques, retrieval-based time-series data analysis is not premised on static mathematical models and does not require extensive computations, so it significantly minimizes the problems involved in fine-tuning the model. By applying this technology to system monitoring, for example, you can automate the assessment normally performed by human observers — i.e., whether the current system condition is the same as usual and doesn’t have any problems, or it is different from usual, or it resembles a particular past event. This paper examines the features of this technology and highlights a number of use cases, including operation monitoring.

NEC is now working on new artificial intelligence (AI) technology for industrial plant operation support that combines logical reasoning with knowledge, enabling it to qualitatively infer how the plant operates by drawing on information contained in operating manuals and design information. Reinforcement learning that incorporates a plant simulator that is used to train the system, making it possible for it to learn optimal operation of a complex plant in a realistic time frame. This paper provides a general overview of this technology and examines the validation results produced using a chemical plant simulator.

The recent emergence of deep learning has brought significant improvements in the accuracy of pattern recognition technologies including that of image recognition. Although training a large amount of data is required in order to achieve a high accuracy, the preparation of such large data amounts is often difficult in applications to real problems. Issues in how to improve accuracy with limited amounts of data are thereby created. This paper introduces two technologies developed for the effective deep learning of a small amount of training data. One is the layer-wise adaptive regularization” method that sets the regularization strength. This varies depending on the layer, according to the structure of the deep-layer network (a deep neural network). The other method is the “adversarial feature generation” that performs training in the middle layers by generating hard-to-recognize features. This paper demonstrates their validity through experiments on the public datasets for handwritten digit recognition (MNIST) and general object recognition (CIFAR-10).

AI needs an extremely high computing performance due to an increase in data scale and complications in algorithms.
Consequently it has become critical to use hardware accelerators arranged for specific purposes. Considering the difficulty for individuals to develop AI that covers a very wide range, the software for supporting the accelerators, called the framework software, has also become important. This paper is intended to introduce Flovedis, a framework for statistical machine learning using supporting accelerators developed by NEC as well as the Vector Engine, an accelerator supporting both statistical machine learning and deep learning.