Advances in geospatial technologies have enabled us to capture an overwhelming diversity of data types. Our focus is to take these data types and develop tools and software that can store, process, analyse and visualise various types of geospatial big data using advanced programming languages to describe the various aspects of the Earth environment.
We aim to act as a pathway between traditional computer science research with spatial-based research, developing state-of-the-art models for various environmental and earth science applications with a key focus on the recent advancement in neural networks. The research outputs have immense societal benefits across various fields and at spatial scales.
One of our core goals in this area is related to Explainable AI (XAI) being used for purposes including earthquake damage mapping, flood susceptibility, landslide susceptibility analysis and more.
Our methods include integrated and ensemble models (using statistical, probabilistic, data mining and artificial intelligence etc.) that are best suited for enhancing the prediction ability of the existing methods in various Earth observation and natural resource management.
Products we developed are designed to be ‘user friendly’ and ‘freely and readily available’ to enable rapid adoption by end users.
