As a demonstration of big data assimilation to predict sudden downpours, this article describes data assimilation in leading meteorological research and provides its broader perspectives beyond meteorology. For the readership of “Applied Physics,” this article focuses on a concrete methodology for data assimilation from the viewpoint of complex models and applied physics. Data assimilation links large-scale complex models and actual data and provides a general methodology for syncing mathematical models and the real world. This article describes the background of numerical weather prediction, the role of data assimilation, and the perspectives to broaden applicability in various fields.

Macro-economic phenomena have remarkable patterns. Such patterns have statistical features that are widely observed in complex systems and unlike physical laws, they can be broken or change under abnormal circumstances. This manuscript focuses on one of the striking phenomena, namely the statistical life and death of firms that are the engine of the real economy. Being helped by recently available large data, including a production network comprising a million firms as nodes and their suppliers-customers relationship as links, and also by large-scale simulations on supercomputers, researchers have been and are finding that concepts and methods in statistical physics are useful to understand social and economic phenomena on a national and global scale.

With the widespread use of smart devices such as smart phones, there are many opportunities to measure the location of people or things, and utilize the location for various applications. This paper outlines technologies for understanding and predicting the flow of people in cities based on the population statistics data that is expected to be used to comprehensively handle the location of people in cities while protecting privacy-related location information.

The creation of artwork is a novel application field for AI. There are studies in the domains of paintings, music and literature. One direction of this trend that we are seeing is projects trying to create artworks that vividly remind us of artists no longer with us, by machine learning of the artists’ style. Following the pioneering attempts regarding the Dutch painter Rembrandt and the Japanese singer Hibari Misora, the TEZUKA2020 project was recently carried out to create a new piece of manga by the legendary Japanese manga artist Osamu Tezuka, who died in 1989. Unlike its predecessors, AI was not intended to produce the final product in TEZUKA2020. Rather, AI was supplying the drafts of plots and character designs with a lot of Osamu Tezuka’s flavor to human manga artists to stimulate their imagination. The unique scheme of manga production has made the approach possible. In this article, the author reviews the character generation process and examines the project as a case study of AI-based creativity support.

Electronic data from Electronic Medical Record systems in hospitals have been accumulated for some time. We are promoting Medical Data Mining research targeting these data. It is expected that these studies will provide new medical evidence (Real World Evidence) from actual clinical data (Real World Data). This paper focuses on the activities carried out at the Center of Medical Information Science at Kochi Medical School, and introduces an approach necessary for the task of converting from actual clinical data to data that can be used for medical research and the construction of an analysis environment. In addition, we will introduce the viewpoints and necessary methods that should be noted when using clinical data for medical research.

The digital transformation of society and business has required higher performance from computing for more advanced and complex large-scale data analysis. The slowdown of Moore’s law increases the expectations for domain-oriented and quantum computing. Digital Annealer, which is a quantum-inspired digital technology architecture that can rapidly solve combinatorial optimization problems, is one of these domain-oriented computers. In this article, we describe the third generation Digital Annealer, and show the practical use of Digital Annealer in the chemical and pharmaceutical fields through introducing recent research examples of molecular structure similarities, battery material development, and drug discovery.

The possibility of a neural network structure utilizing any response from devices, materials or physical substances, for which a function can be used for information processing, is now attracting considerable attention. The realization of a neural network consisting of materials requires experimental processes and knowledge that might not be common in the research field of applied physics. We demonstrated the structure of two different types of neural networks utilizing the plasticity of polymer growth and the nonlinear response of the electrochemical reaction of molecules. Our organic neural networks show primitive functionality for information processing. In this paper, the system of basic neural network and reservoir computing are illustrated and explained in plain words, and experimental processes, namely, how the materials learn to be a neural network and how we evaluate their performance, are concretely explained.

Rechargeable batteries have been used in various fields ranging from portable devices to electric vehicles. To improve rechargeable batteries, a vast amount of research will need to be conducted. In this text, the fundamental points of battery research are shown and then cyclic voltammetry and charge-discharge measurement are chosen for the analysis of battery reactions, in particular, for graphite negative electrodes for lithium-ion batteries.