Static, citable disaster-risk explainer for Japan's PLATEAU open data. Building-level hazard zones rendered with strict three-state honesty (covered=false is never green).

This is the repository of the Webapp, that is a product of Tomorrow's Cities Project, which aims at simulating future urban context and modelling potential impacts against multi-hazards.

BARLO: Bayani Alert and Response for Local Operations. Predicts a storm's economic impact from typhoon forecast data using a PyTorch + scikit-learn model, deployed on Streamlit.

Malaysia-focused flood risk prediction and explainability research project using FastAPI, Streamlit, geospatial data, and machine learning.

CCART is a country-agnostic physical climate risk framework. India is the reference implementation. The architecture is designed so that any country can be added by providing local boundary files and climate data

個人初の長期開発「リアルタイム災害脆弱性ダッシュボード」。

Open earthquake digital twin for Greater Dhaka — hazard, exposure, cascade, surrogate emulator

[8th IDRC, Best Overall Conference Paper Award] GraphCSVAE: Graph Categorical Structured Variational Autoencoder for Spatiotemporal Auditing of Physical Vulnerability Towards Sustainable Post-Disaster Risk Reduction

Philippines Storm Impact Analyzer: a Streamlit web app combining typhoon forecast data with geographic and economic data to estimate damage and affected provinces. Built for the PJDSC competition.

CRIMA ontology source files