About the Project
This project develops novel AI methods for property risk assessment under climate and disaster-related hazards, addressing challenges of fragmented data, uncertainty, and passive data use. It investigates uncertainty-aware multimodal data fusion and decision-theoretic data acquisition strategies to reduce uncertainty in risk estimation. By combining distributed digital data sources with satellite observations and leveraging computer vision techniques, the research enables adaptive and dynamically updateable risk assessment. A scoped predictive case study demonstrates the value of these methods. Co-designed with industry, the project supports improved decision-making for insurers and policymakers, contributing to climate change adaptation and disaster resilience.
Name of primary supervisor/CDT lead:
Name of secondary supervisor:
Huili Chen
Bench fees required: No
Closing date of advert: 9th June 2026
Start date: October 2026
Full-time/part-time availability: Full-time 3.5 years, Part-time 7 years
Who is eligible to apply?: Both UK and International