Funded by Microsoft, the grant will support high performance computing through $150,000 in Azure computing credits. The collaboration between the Texas A&M Institute of Data Science (TAMIDS) and the College of Architecture will use AI to leverage social media big data, such as text, images, and videos to enable rapid damage assessment for disasters.
Since 2018, Microsoft has supported 42 AI for Humanitarian Action projects in 17 countries, advancing solutions to address challenges in disaster response, refugees, displaced people, human rights, and the needs of women and children through grants, technology donations, and data science support.
Dr. Yalong Pi, Principal Investigator on the project and Postdoctoral Data Scientist in TAMIDS Operational Data Science Lab states: “Disasters of natural origin have severely impacted our society and extreme weather has become more frequent under the changing climate. Economic damage assessments are crucial for allocating resources for recovery rapidly where they are needed. However, damage estimation currently relies on self-reporting, on-site surveys, and insurance claims, all of which have a latency of weeks to provide information to guide efficient distribution of recovery resources. For example, loss estimation of the 2021 Winter Storm in Texas is still not widely available after 4 months. The challenge that we want to address is to provide damage prediction and loss estimation in a timely manner by incorporating social media data”.
Working alongside Dr. Pi will be Co-PI’s Dr. Xinyue Ye, Associate Professor and Harold Adams Endowed Professor, Department of Landscape Architecture and Urban Planning, College or Architecture and Dr. Nick Duffield, TAMIDS Director and Royce E. Wisenbaker Professor I in the Department of Electrical and Computer Engineering, College of Engineering. Dr. Jisu Cao, formerly a postdoctoral researcher with the Mays Business School and TAMIDS, also contributed to the proposal.