Itoi Lab
Promote natural disaster risk management by integrating science, technology and society to realize an evolving society that continues learning from disaster risk.
Research Theme
To contribute to the development of natural disaster risk management, we are conducting research on performance evaluation of buildings, urban risk assessment, prediction, and evaluation of natural disasters such as earthquakes and tsunamis. We approach these research issues using methods such as simulation based on analytical models, probabilities and statistics, machine learning utilizing big data / ICT, and propose a new framework related to natural disaster risk assessment.
Performance Evaluation of Buildings
By utilizing observed seismic records, we are developing a technology to clarify the structural performance of buildings and to predict structural damage in future earthquakes. We also evaluate the performance of indoors and non-structural elements by applying machine learning and deep learning using image and video data. Based on this research, we will evaluate the performance of buildings in terms of life and business continuity after a disaster. In addition, we are developing a methodology to provide information for design decision making by optimizing the lifecycle costs of buildings using big data of seismic motion observation records.
Urban Disaster Risk Assessment
We are developing methods to evaluate the damage caused by future earthquakes and tsunamis at the urban scale. We are conducting research to monitor the urban performance from various perspectives, such as a building or a group of buildings in a city. We will evaluate urban performance of buildings, infrastructure, etc. through machine learning using various sensing data. By proposing a new urban resilience framework through these studies, we will contribute to realize a disaster-resistant society.
Seismic Motion Prediction
We are conducting research focusing on earthquake hazard themselves to be used for building structural design and disaster risk assessment. By evaluating the ground characteristics using seismic motion observation records, we will realize more accurate seismic motion prediction. We are also developing a novel data-driven approach to predict earthquake ground motion based on artificial intelligence learned from big data of earthquake records.
