Electronics and Electrical Engineering
Research Theme
Aim
To develop a combined radar cross-section and synthetic aperture radar machine learning based classifier.
Objectives
1. Adopt data fusion approaches combining RCS and ISAR information for the studied targets and for the various polarization cases with machine-learning (ML) based methods for target classification.
2. Assess classifier robustness with respect to aspect angle, noise levels, data sampling variations, target positional information, and potential reduced information scenarios.
3. Develop a preliminary data set using simulation, polarization-specific, and measured data of relevant targets while using measurement and modelling facilities within The Scottish Microelectronics Centre (SMC).
Description
There is an increasing defense requirement for classification techniques for airborne, terrestrial, and seaborne targets. Typical radar cross-section (RCS) signatures provides information on the electromagnetic scattering of targets while synthetic aperture radar (SAR) provides a two-dimensional representation of the distribution of scattering features in terms of 2D images. These two forms of information are complementary yet different: RCS signatures and measurements describe the aggregate scattering response under particular observation conditions and polarization states, whereas SAR can provide relative spatial details on the scattering features of targets.
The PhD research work will aim to investigate a machine-learning (ML) framework for classifying a broad range of airborne, terrestrial, and sea-borne targets using both RCS signatures and ISAR representations. The main motivation will be to assess if non-imaging-type RCS signatures along with SAR-derived information, can provide more robust classification of targets, and compared to when these representations are independently examined. Numerical modelling, full-wave simulations using commercial software tools, and measurement information strategies will be adopted while using the existing facilities within the RF lab at UoE.
An engineering or physics background is suitable, but engineering is the preference.
Minimum entry qualification - an Honours degree at 2:1 or above (or International equivalent) in a relevant science or engineering discipline, possibly supported by an MSc Degree.
This position is fully funded.
Research Themes
Aim
Develop an integrated probabilistic decision-support system for detecting and evaluating the origin of nefarious microbial threats.
Objectives
- Fusion of diverse data sources (e.g. observations, biological, genetic) into an integrated system.
- Modelling and propagation of uncertainties, from quantitative data (e.g. measurement uncertainty), subjective information (e.g. intelligence data) and uncertainties in the possible origin or dissemination of pathogenic microbes.
- Development of a fast, usable and generalisable graphical decision support system for supporting real-time and uncertainty-aware decisions.
Description
The aim of this project is to integrate Bayesian probabilistic/AI graphical modelling approaches with diverse datasets to assess and characterise microbial threats - combining subjective and background information, genomic data, proteomic and toxin data, and microbial growth data to provide a comprehensive and dynamic view of pathogenic microbial encounters. The research will develop probabilistic methods for distinguishing between natural, intentional or accidental outbreaks of pathogenic microbes – this is crucial for understanding the current threat from bioweapons.
In some cases, the question of nefarious intent might be answered almost definitively by a specific type of omics-based science, such as in the case where very clear genetic manipulations are evident in a microbe’s DNA. However, in many other cases, the proposition of nefarious intent will need to be assessed in probabilistic terms and there will be a requirement to incorporate a diverse suite of non-definitive but corroborating information in one model. Bayesian graphical modelling approaches such as Bayesian Networks and Chain Event Graphs provide an appropriate framework for integrating this information and have been used successfully in a forensic science context.
The challenges to be addressed by the project include applying these new approaches to real-world decision-making processes, adapting models for different pathogens in real-time investigative situations, and incorporating complex and highly diverse biological and contextual information.
Some background and interest in statistics or probability, interest in applying mathematical methods to biology.
Minimum entry qualification - an Honours degree at 2:1 or above (or International equivalent) in a relevant science or engineering discipline, possibly supported by an MSc Degree.
This position is fully funded.
Research Themes
Multi-agent Systems and Data Intelligence
Aim
Develop statistical learning methodology for active sonar that provide robust detection and classification across noisy underwater environments while producing calibrated uncertainty estimates to support reliable defence and security decision-making.
Objectives
The project is divided into an initial exploration stage, followed by one of three possible research avenues, with various possible extensions.
- Exploration. Through a review of existing literature, available data sources, and exploratory numerical experiments, investigate the most promising route for maximising the potential of statistical learning methodology in classifying active sonar data while retaining useful quantification of uncertainty.
- Methodology development. Enable transfer across frequency bands, sensor configurations, environments, and target classes, including adaptation from broad pretraining to specific operational detection tasks. Depending on the exploration stage, this may involve
- the development of a foundation model for active sonar data which can be fine-tuned to specific domains,
- a transfer learning model to exploit knowledge from one active sonar domain with large amounts of available data (such as high-frequency side scan sonar) to one with little labelled data (low frequency sonar),
- an active learning approach in which the model works only on the target domain data, first clustering it and building a statistical classifier based on human prioritisation of the detected clusters.
- Enable practically useful and useable uncertainty quantification, e.g. through Bayesian deep learning approaches or conformal prediction, focusing on uncertainty-aware classification and detection methods in high-clutter situations so that model outputs provide operationally meaningful confidence estimates and prediction sets.
Description
Active sonar is a key sensing technology for underwater defence and security, but reliable detection and classification remain challenging as acoustic returns are strongly affected by propagation conditions, reverberation, environmental variability, target aspect, and low signal-to-noise ratios. Labelled active sonar data are limited and often highly scenario-specific, making it difficult to train models that generalise reliably to new environments or sensor configurations.
Existing machine learning research in the acoustics domain focusses largely on passive monitoring due to the financial costs associated with both employing active sonar and labelling collected data.
This project will investigate statistical learning approaches for active sonar classification. The student will develop machine learning methods that learn transferable representations from large collections of underwater acoustic data, using self-supervised, weakly supervised, or simulation-assisted training where appropriate; the approach will be determined based on initial experiments, a literature review, and available data sources.
A central focus will be uncertainty quantification. The project will study methods such as Bayesian neural networks, deep ensembles, conformal prediction, calibration techniques, and probabilistic classifiers to ensure that the system outputs not only a prediction, but also an operationally meaningful confidence measure.
The probabilistic outputs of the methodology will be used in a downstream task in a practically useful way, such as flagging uncertain detections to a human operator.
A strong mathematical background and programming experience, preferably in Python, are desirable. Prior experience in machine learning, signal processing, Bayesian statistics, acoustics, or underwater sensing would be advantageous but is not essential.
Minimum entry qualification - an Honours degree at 2:1 or above (or International equivalent) in a relevant science or engineering discipline, possibly supported by an MSc Degree.
This position is fully funded.
Developing renewable energy solutions that can be rapidly implemented in the market using eco-friendly materials and manufacturing methods is crucial. Among various renewable technologies, photovoltaics have significant potential to support climate change mitigation. Organic photovoltaics (OPVs) have recently attracted considerable attention due to a new family of semiconductors that enable highly efficient light harvesting in both indoor and outdoor environments. Additionally, OPVs offer a low carbon footprint and high recyclability potential.
However, a current limitation is the use of toxic solvents and materials in manufacturing. Most organic electronic devices require halogenated and non-halogenated aromatic solvents, which are often carcinogenic or toxic to human reproductive systems and harm the environment. For large-scale production and commercialization, this is a critical issue.
This project aims to enhance the performance of OPVs through engineering strategies, eliminate the use of toxic materials and implement methods to enhance their stability. In addition, thin films and OPVs will be evaluated with a series of optoelectronic and morphological characterisation tools.
The PhD candidate will be supervised by Dr Julianna Panidi (School of Engineering) and Dr Yue Hu (School of Chemistry).
The successful candidates will join our team, which includes researchers from the Institute for Integrated Micro and Nano Systems, the School of Chemistry, and the wider College of Science and Engineering.
Eligibility
- a 2:1 undergraduate degree (or equivalent).
- applicant's must meet the University’s English language requirements.
Before you apply: We strongly recommend that you contact the supervisor for this project before you apply.
https://eng.ed.ac.uk/sustainable-printable-electronics-research-group
Applications are welcomed from self-funded students, or students who are applying for scholarships from the University of Edinburgh or elsewhere.
Funding may be available to Home students on a competitive basis. Candidates are invited to discuss potential funding opportunities.