Vol. 76, No. 5, September 2021

SHIMADA Taro

OMICHI Shigeo

Attention is being focused on managed services as a type of outsourcing service evolved from information technology (IT). Managed services enable users to concentrate on their business operations in response to the rapidly changing requirements of the market by utilizing a variety of asset-light services.

The Toshiba Group has set the goal of becoming a cyber-physical systems (CPS) company capable of delivering efficient infrastructure services. We are making efforts to create and provide new value to customers and society as a whole through the enhancement of managed services based on our tangible and intangible assets, including knowledge and proprietary technologies for systems and components that we have accumulated through the development of infrastructure systems.

IKETANI Naoki / KANO Ichiro

In line with the recent trend toward the introduction of cloud managed services aimed at realizing cyber-physical systems (CPS), various companies have found it necessary to replace the individual management of multiple cloud services designed for specific purposes with unified management common to all cloud services, in order to further improve operating efficiency and ensure higher security.

Toshiba Digital Solutions Corporation has responded to this need by developing a framework for a virtual private data center as a solution enabling the operation of multiple cloud services through the standardization of security, network, monitoring, operation technologies. This virtual private data center makes it possible to swiftly construct a unified cloud system incorporating the user’s private cloud systems while also reducing operating costs.

HAYASHI Takanori / HIRAYAMA Tsuyoshi / ASHIKAWA Masayuki

Attention has become increasingly focused on artificial intelligence (AI) in a broad range of fields. However, the high cost of introduction and difficulties involved in continuous operation are critical issues that need to be overcome in order to achieve the practical use of AI technologies.

To rectify this situation, Toshiba Digital Solutions Corporation has developed and is operating the SATLYS AI Common Platform to provide elemental technologies for AI services as a managed service featuring user-friendly functions. This platform provides comprehensive support for processes from the preparation of teaching data through to the monitoring of model accuracy, making it possible to achieve the rapid construction and continuous operation of AI services in response to customers’ requirements.

FUKUSHIMA Nobuyuki / KONDO Yuji / FUJITA Shinichi

Cloud managed services are expected to substantially save labor and reduce costs in the operation and maintenance of information technology (IT) systems including servers, as well as to handle failures occurring in IT systems more effectively, compared with on-premises environments.

Toshiba Digital Solutions Corporation has already developed and released on-premises GridDB scale-out databases for various Internet of Things (IoT) systems. We have now developed and launched GridDB Cloud, a cloud-based database service with features inherited from the GridDB scale-out databases. GridDB Cloud makes it possible to reduce maintenance costs and shorten the database introduction period through IT outsourcing.

KUMAZAKI Yuichiro / IMAI So

Managed services for the maintenance of computer systems, which can effectively analyze large volumes of data, have been attracting increasing attention in recent years in line with the expanding number of monitored devices accompanying the dissemination of the Internet of Things (IoT).

With this trend as a background, Toshiba IT-Services Corporation has been providing managed services based on its know-how acquired through the accumulated experience of its engineers. In the field of remote monitoring services, we are contributing to the stable operation of customers’ systems by providing an incident management system that automatically classifies alarms and extracts critical incidents needing specific actions for system recovery. Our incident management system has achieved a 96% reduction in the number of manual operations required for such incidents.

ASAHI Kazutaka / FUCHIKAMI Yasunori / SATO Masataka

In the logistics business field, there is an urgent need for labor-saving and further automation of postal processes to handle the increasing volume of parcels being generated by e-commerce transactions as a consequence of the worldwide spread of COVID-19. On the other hand, parcel sorting machines equipped with conventional address optical character recognition (OCR) systems to read address information, including both address details and postal-code data, are showing an increasing tendency to generate errors accompanying the growth in the number of mail items of diverse types and shapes.

As a solution to these issues, the Toshiba Group has developed a dedicated OCR function focusing on the recognition of address information that parcel sorting machines cannot conventionally recognize in address images, and is providing it as a pay-per-use secondary OCR service for users of existing parcel sorting machines. We have confirmed that this secondary service can successfully read about 70% of unrecognized address images under certain conditions. With the objective of expanding this service, we are also developing a new secondary OCR system that can be easily connected to existing parcel sorting machines through the implementation of a manipulation technique.

SETOGUCHI Tatsuya

Swift management decisions are required in response to changes in the business environment due to unexpected events including disasters and incidents, which have become increasingly difficult to predict in recent years. This is particularly true for the procurement departments of manufacturing industries, making it essential for them to plan and execute their procurement strategies in a flexible manner.

In this context, Toshiba Digital Solutions Corporation has developed Meister SRM (SRM: Supplier Relationship Management), a strategic procurement solution aimed at responding to customers’ changing procurement strategies, and has been providing it in the form of software as a service (SaaS) applications deployed via a cloud management service based on Microsoft Azure. In order to continuously support customers’ strategic procurement needs, we have now further enhanced the service functions of Meister SRM by systematizing the processes for its introduction, establishing a methodology to provide secure services within a short time, and facilitating collaboration with a variety of external cloud services.

HAMAKAWA Yohei / HIDAKA Ryo / TATSUMURA Kosuke

The Toshiba Group has developed a simulated bifurcation algorithm based on its proprietary quantum-inspired algorithms to rapidly solve combinatorial optimization problems that frequently arise in social and industrial settings.

In addition to the Simulated Bifurcation Machine (SBM) for cloud services released in 2019, we have now developed SBM for on-premises services, which can be embedded in real-time systems. The on-premises service is being provided to users as a packaged module consisting of configuration data to realize a high-performance processing circuit on a field-programmable gate array (FPGA) and an application programming interface (API) with simplicity and versatility, in addition to newly developed reference designs to verify applicability to real-time systems. This service is a focus of high expectations as a real-time Ising machine for the expansion of applications in various fields and the co-creation of new value.

OKANO Motochika / IMI Hitoshi / OKURA Gentaro

In line with the recent trend toward the electrification of automobiles, demand has been increasing in the automotive field for power semiconductors as key devices of in-vehicle systems. With the increasing use of electronic components, the introduction of simulation technologies using model-based development (MBD) methods to verify the functions and performance of overall systems has been rapidly expanding in order to efficiently develop in-vehicle systems.

Toshiba Electronic Devices & Storage Corporation has been promoting the deployment of power semiconductor products including metal-oxide-semiconductor field-effect transistors (MOSFETs) in response to customers’ requirements for in-vehicle systems, and supplying not only devices but also device models of MOSFETs for such simulations. Moreover, we have been carrying out simulations of in-vehicle systems using our device models with the objective of offering optimal device models for system-level simulations. We found during this work that it was necessary to resolve the tradeoff between accuracy and calculation time. To achieve a balance between high accuracy and high speed, we have now developed the Accu-ROM (accurate reduced-order modeling) technique focusing on the evaluation of thermal performance and electromagnetic interference (EMI) noise in switching operations and confirmed the effectiveness of this approach using the Accu-ROM technique when selecting an optimal MOSFET device and switching speed.