Fujitsu and RIKEN Jointly Open the "RIKEN AIP-FUJITSU Collaboration Center" for Developing Human-Centric AI

Fujitsu Is Aiming for Human-Centric AI

In recent years, along with globalization and digitization, social and economic systems have become diversified and complicated. It is almost impossible for humans as decision makers to clarify the scope of all social changes and take appropriate strategic measures. However, for social issues that are difficult to surmount, such as an aging population combined with the diminishing number of children and natural disasters, it is required to predict the future appropriately and make better judgments.

Under such circumstances, utilizing artificial intelligence (AI) is attracting increased attention these days. AI has the potential to dynamically transform industry and society by speedily supporting human judgments and activities.

Fujitsu has a vision to develop "human-centered AI that collaborates with people," with the aim of realizing the fulfillment of lives through the utilization of AI, putting people at the center. Fujitsu structured this human-centered AI as "Human Centric AI" and is aiming to solve issues by combining people and its automatically evolving AI, and has also begun incorporating it into a variety of products and services.

Collaborating with RIKEN AIP and Appropriately Predicting the Future

The RIKEN Center for Advanced Intelligence Project (AIP Center) and Fujitsu have recently opened the "RIKEN AIP-FUJITSU Collaboration Center" and started operations. By integrating AI technologies collected at the RIKEN AIP Center and Fujitsu's wide variety of ICT technologies and system development experiences, Fujitsu will work on R&D to create "AI that predicts the unpredictable."

"AI that predicts the unpredictable" is AI that supports better decision-making for people based on appropriate future predictions, even in the face of environmental uncertainties. Toward this realization, the RIKEN AIP-FUJITSU Collaboration Center will advance research on Robust Machine Learning, Fusing Simulation and AI, and Structuring Large-Scale Knowledge, forming a trinity.

For example, conventional machine learning has a fundamental issue in that, without massive volumes of high quality data, it cannot provide sufficient predictive capabilities. In order to overcome this issue, the center aims to appropriately predict the future, even with limited amounts of or incomplete data.

This AI technology is expected to be applied to many fields, with the following three fields as examples:

(1) Social Infrastructure
Ensuring both appropriate defense and continued operation of internal systems by not only detecting attacks but also anticipating the range of influence when a new type of virus is detected.

(2) Healthcare
It is said that a change occurs in the brain 10 years before the onset of dementia. AI that can detect a small change in the brain makes it possible to discover such subtle changes at an early stage to help prevent symptoms of dementia. It can also be applied to situations that are difficult assess, such as intractable diseases, rare diseases and new kinds of influenza.

(3) Manufacturing
AI can be utilized to discover unexpected new materials beyond the evolution of semiconductors, aiming to achieve "More than Moore" *.

*: The idea of creating new functionality by integrating semiconductor elements having different materials and functions into one, outgrowing Moore's Law, which states the integration density of semiconductor chips will double everyone-and-a-half to two years.

Toward the Creation of "Japan First" and "World First" Innovations

Other than the above, Fujitsu and RIKEN will closely collaborate in the AI field, such as building the "Deep Learning System." Fujitsu and RIKEN will continue striving to create "Japan First" and "World First" innovation by conducting consistent research; from the development of basic technologies to their eventual implementation in society.