Ordnance Survey (OS) own and manage diverse sets of national geospatial gridded and vector data alongside records about addresses, network topologies and the heights of the terrain and buildings. These datasets are used to address national-scale issues both in Great Britain and internationally, such as mapping land use or estimating greenhouse gas emissions and drawdown. However, these datasets are prohibitively large and contain redundant information.
The goal of this PhD is to develop a new framework for reducing the volume but not the useful information in the data. Using statistical models and machine learning, methods should have the ability to do this using the data itself in addition to allowing input in terms of prior knowledge of the scene such as geographic location, urbanisation, and season. The main purpose of this framework is the development of a method of automatically curating OS imagery data sets such that they result in information equivalent training outcomes with a smaller volume of data.
Challenges in this project include the processing and combination of large geospatial datasets which have different levels of support, have different levels of aggregation, are spatially misaligned and have their own sources of biases and errors. Statistical inference in large spatio-temporal data will require advanced and modern scalable computational methods, using techniques such as Approximate Bayesian Computation. The definition of “useful information” may be different from one application or training scenario to another and this will need to be understood and accommodated.
The successful applicant will study existing approaches to similar problems, while developing a deep understanding of the data inside and outside of OS’s extensive catalogue. They will work closely with OS to develop new methodology to handle and integrate large and diverse sets of data in a mathematically coherent and a computationally efficient way. The methods developed will be applied to curate OS’s imagery data such that they result in information-equivalent outcomes with a smaller volume of input data. Ultimately, the methodology will be validated with focus on the operational decision making for OS.
The studentship is funded via the EPSRC Industrial Cooperative Awards in Science & Technology (CASE) scheme, via a grant awarded to OS. The successful applicant will be required to undertake regular reporting and visits to the sponsor. For more details about the Industrial CASE scheme, see https://epsrc.ukri.org/skills/students/industrial-case/intro/
The successful applicant will be aligned to the UKRI CDT in Environmental Intelligence and will be included in CDT cohort building and training activities where suitable. Unlike standard CDT students, the successful applicant will be attached to the above project from the beginning of their PhD under the supervision of Dr. Matthew Thomas and Prof. Gavin Shaddick. They will not be permitted to change to a different project.
About the UKRI Centre for Doctoral Training in Environmental Intelligence:
Our changing environment presents a series of inter-related challenges that will affect everyone’s future health, safety, and prosperity. Environmental Intelligence (EI) is the integration of environmental and sustainability research with data science, artificial intelligence, and cutting-edge digital technologies to provide the meaningful insight to address these challenges and mitigate the effects of environmental change. One of the 16 UKRI AI CDTs launched in 2019, the CDT in Environmental Intelligence provides an interdisciplinary training programme for students covering the range of skills required to become a leader in EI:
· the computational skills required to analyse data from a wide variety of sources,
· Expertise in environmental challenges,
· an understanding of the governance, ethics and the potential societal impacts of collecting, mining, sharing and interpreting data, together with the ability to communicate and engage with a diverse range of stakeholders.
The CDT cohort (currently around 30 students out of the planned 50), works and learns together, bringing knowledge, skills, and interests from a range of academic disciplines relevant to EI. CDT students undertake training and professional development as a cohort, and regularly participate in seminars, symposia, and partner engagement activities including the annual CDT Environmental Intelligence Grand Challenge. As part of the research community at the University of Exeter, CDT students benefit from networking with colleagues in the Institute for Data Science and Artificial Intelligence; the Global Systems Institute; and the Environment and Sustainability Institute.
This studentship will cover home tuition fees plus an enhanced annual tax-free stipend of £20,062 for 4 years full-time.
Applicants for this studentship must have obtained, or be about to obtain, a First or Upper Second Class UK Honours degree, or the equivalent qualifications gained outside the UK, in an appropriate area of statistics, mathematics, data science or computer science. science or technology.
English language requirements - (Profile A) at https://www.exeter.ac.uk/pg-research/apply/english/
How to apply
In the application process you will be asked to upload several documents:
· Letter of application (outlining your academic interests, prior research experience and reasons for wishing to undertake the project).
· Transcript(s) giving full details of subjects studied and grades/marks obtained (this should be an interim transcript if you are still studying).
· Two references from referees familiar with your academic work. If your referees prefer, they can email the reference direct to [Email Address Removed] quoting the studentship reference number.
· If you are not a national of a majority English-speaking country, you will need to submit evidence of your proficiency in English.
Closing date for applications is midnight Friday 9th September 2022.