Research conducted in the Dr. Takuya Oki Laboratory

Abstract

Associate Professor Takuya Oki, Tokyo Institute of Technology, Japan, is using big data and AI technology to address issues associated with architectural planning and urban planning research. Through his research he hopes to provide basic data that can be used in architectural and urban design and in the design and development of urban spaces. He also predicts his research can be used in the analysis of problems within existing spaces and the provision of improvements. Oki is also working to develop a real-time evacuee distribution estimation technology in collaboration with the University of Tokyo's Assistant Professor Yoshiki Ogawa and Associate Professor Masaaki Imaizumi, and the Shibaura Institute of Technology's Associate Professor Yuki Oyama. The goal of this work is to help solve the problems of developing a wide-area evacuation simulation model that was originally proposed in Oki's doctoral thesis. To do this, the researchers are using an AI technology called a graph neural network (GNN) model to learn the movement tendency of evacuees from a limited number of simulation results. The team is also using data assimilation to reduce error. Oki is also working on research for the Consortium for Socio-Functional Continuity Technology (SOFTech) surrounding 'ensuring peace of mind', which involves analysing the gazing tendencies of TV viewers and pedestrians when disaster prevention-related information is presented.