Two funded PhD Studentships now available in the IRDR
9 January 2023
Applications are now open for two studentships in the IRDR, as part of 60 studentships to be awarded by the UCL EPSRC DTP.
The UCL EPSRC DTP (Doctoral Training Partnership) has 60 fully-funded four-year studentships available for 2023/24. Prospective students may apply to any of the projects available in the project catalogue but there are two based in the IRDR.
Direct Multi-Hazard Consequence-Based Design of Buildings
Lead project supervisor: Roberto Gentile
Project ID: 2228bd1174 (you will need this ID for your application)
In 2021, natural hazards caused 10,500 deaths and 252B US$ losses. Such numbers will increase in the future due to (often-uncontrolled) urbanisation and exacerbation of climate-related hazards. Most of this future risk involves yet-to-be-built assets, among which residential buildings constitute a relevant share. There is an opportunity to use risk-targeted design approaches to invert the above trend.
Performance-based earthquake engineering allows calculating consequences (e.g., deaths, dollars, downtime) for building configurations (including structural and non-structural components), using time-consuming numerical analyses involving several ground-motion excitations. Since this non-linear assessment formula cannot be inverted, it is used iteratively to design new buildings (i.e., obtaining a building configuration consistent with a target level of consequence). Using machine learning and surrogate modelling, the supervisor developed “Direct Loss Based Design (DLBD)”, which allows designing buildings to achieve a target economic loss for a site-specific seismic hazard. The procedure is “direct” because it does not require design iterations, thus enabling design in the engineering practice. DLBD is currently limited to earthquakes, and it can only target economic losses.
The candidate will develop DLBD for flooding using conventional (e.g., economic loss) and people-centred (e.g., human displacement) consequence metrics. This involves: a review of the state-of-the-art flood risk assessment of buildings; a sensitivity analysis to identify the most influential parameters; machine-learning techniques to surrogate the building configuration-to-consequence mapping (the core of flood DLBD). To include climate change effects on the flood severity/yearly rate during the building design lifetime, a lifecycle approach will be used. Finally, flood DLBD will be combined with the existing seismic DLBD to obtain “direct multi-hazard consequence-based design”.
Who we are looking for:
The ideal candidate for this project has a master’s degree in earthquake engineering and a reasonable coding level. Proficiency in probabilistic risk analysis is desirable but not essential.
Digital Twins for Disaster Early Warning Systems
Project supervisors: Prof Ilan Kelman and Dr Saman Ghaffarian
Project ID: 2228bd1131 (you will need this ID for your application)
The main objective of this project is to conceptualise, design and develop digital twins-based disaster early warning systems that dynamically learn from (near) real-time data and predict impacts for effective and timely early actions.
Early warning systems as a crucial part of disaster risk management aim at enabling vulnerable communities or systems to take timely actions and mitigate disaster risks in advance of the imminent hazardous event/s. Current early warning systems are based on conventional rule-based modelling usually focusing on one or a few components and elements at risk, probabilistic scenarios and impact assessments largely using static predictions and manually derived information. However, such systems require complex and dynamic modelling. A Digital Twin is a digital equivalent to a real-life object, process, or system of which it mirrors its behaviour and states over its lifetime in a virtual space. Using Digital Twins as a central means for early warning systems enables decision-makers to manage early warning-based operations from a holistic point of view employing real-time digital information. This allows them to act immediately in case of deviations and to simulate effects of interventions and decisions based on real-life data.
Who we are looking for:
- You have a Master of Science degree, preferably in Computer Science, Geo-Informatics, Disaster Risk Management, Remote Sensing, or related fields.
- You are fluent in python, machine learning, and deep learning tools.
- You have knowledge of disaster and risk reduction.
- Expertise in Big data analytics and/or Internet of Things is an advantage.
Funder: UCL ESPRC DTP studentship
Value: Fees, Stipend (at least £20,668 per year), Research Training Support Grant
Duration: Up to 4 years (thesis to be submitted within funded period)
Eligible Fee Status: Home, International (EPSRC caps the total number of funded International fee status students across UCL for this award at 30%)
Study Mode: Full or Part time (at least 50% FTE) [Note: Part time is not available to International students]
Primary Selection Criteria: Academic merit
Application Deadline: 12:00 GMT on 26 January 2023
How to apply
Visit the UCL EPSRC DTP studentship webpage for more information on how to apply.
You will need to use the relevant Project ID in your application.
Applications close on 26 January 2023.