Image
Imaging, Data and Communications
Research Theme
Aim
Develop multimodal virtual sensing methods that leverage data-driven models to infer rich sensory feedback from minimal physical inputs, thereby enabling adaptive and reliable robotic manipulation in cost-constrained and hazardous environments.
Objectives
- Develop AI-based virtual sensor models that can approximate multimodal feedback (vision, haptics, force-torque, audio) from limited real sensor inputs.
- Design and evaluate multimodal fusion strategies for learning robust object property and interaction representations that generalize across tasks and environments.
- Investigate transfer learning and domain adaptation methods to enable deployment of virtual sensors trained on simulation or rich offline datasets to real-world robotic platforms.
- Validate virtual sensing for manipulation tasks by benchmarking performance against fully instrumented systems in both controlled and hazardous/constrained scenarios.
Description
The PhD project investigates virtual sensing for robotic manipulation, focusing on the use of data-driven models to approximate multimodal sensory feedback. The core objective is to train AI models on rich sensory datasets (e.g., vision, haptics, force-torque, proprioception, audio) to learn robust representations of object properties and interaction dynamics. At deployment, these models will infer missing modalities from minimal physical sensing, enabling reliable manipulation in cost-constrained or hazardous environments. Key research challenges include multimodal fusion, domain adaptation, and the transfer of representations from simulation or offline data to real-world robotic systems. The project aims to advance theoretical understanding of virtual sensing architectures while delivering practical methods for adaptive, resource-efficient robotic manipulation.
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. Further information on English language requirements for EU/Overseas applicants.
Full funding is available for this position.
On
Research Theme
Aim
To develop a defence-ready, low-cost alternative to AI post-training that lets air-gapped AI systems rapidly adapt in the wild
Objectives
- Build a local “module bank” of lightweight adapters (LoRAs) covering defence relevant sensors and tasks.
- Train a learned router that picks and combines the right modules for new scenes, using crossmodal alignment and simple ondevice self-supervision
- Prove the system on air‑gapped edge hardware, measuring speed, accuracy, and reliability under distribution shift
Description
We will develop a low-cost alternative to AI post-training that relies on exploiting a local bank of neural modules, implemented with parameter-efficient adapters (e.g., LoRA) and related lightweight components, for deployment on defence and security platforms. This will allow an AI in the wild—operating in air‑gapped or contested settings—when exposed to a new scenario (e.g. an unidentified object, a change of scene) to rapidly adapt to what it senses using the most appropriate modules, guided by a learned router that selects and composes modules and supports lightweight on‑device self‑supervised retraining to track distribution shift. We will optimise this module bank using both generative and discriminative tasks across defence‑relevant modalities (e.g. natural images, EO/IR, SAR/sonar, RF, audio) and develop algorithms that are able to express a new task/modality in terms of these tasks/modalities to identify the most appropriate modules via cross‑modal alignment into a shared task/sensor embedding space.
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. Further information on English language requirements for EU/Overseas applicants.
Full funding is available for this position
On
Research theme
Multi-agent Systems and Data Intelligence
Aim
To devise novel, explainable, and robust speech deepfake detection methods that go beyond current and next-generation techniques.
Objectives
- Understand the vulnerabilities of State-of-the-Art deepfake detectors when faced with the latest methods for speech synthesis.
- Construct purposely-designed attacks and simulations of beyond-next-generation synthetic speech to further explore these vulnerabilities.
- Devise novel approaches to constructing a deepfake detector that address the identified vulnerabilities, and go beyond binary fake vs. bonafide decisions by providing measures along multiple dimensions.
- Explore the accuracy, utility, and usability of the detector and its explanations, when applied to use-cases in the field.
Description
Digitally created or manipulated synthetic speech of remarkably high naturalness is a reality. Beneficial uses include text-to-speech for people who cannot speak, and privacy-preserving identity protection. But speech deepfakes enable deception, from scam calls to large-scale election interference on social media. Current automatic detection methods learn to rely on specific low-level artefacts in the synthetic speech used to train them, making them vulnerable to newly-emerging deepfakes, whilst incorrectly classifying some natural speech as fake.
In this project, you will rethink deepfake detection. For example, can we take inspiration from human judgments? Humans do not need to have previously heard a specific type of deepfake, presumably because we have a strong internal model of natural speech. Can we decompose “naturalness” into useful dimensions that enable explainable deepfake detection, whether fully-automated or human-AI collaboration (e.g., tools for speech forensic scientists). Better methods of speech evaluation will also advance research in speech generation for beneficial use-cases.
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. Further information on English language requirements for EU/Overseas applicants.
This project combines elements of audio engineering, linguistics, speech science, forensics, and AI, so we are looking for applicants who have a background in one or more of those areas and the aptitude to acquire the necessary skills in the other areas during the PhD.
Note: we do not require an undergraduate degree in a STEM subject but would, for example, consider an applicant with an undergraduate degree in linguistics plus a Masters in Speech & Language Processing.
Full funding is available for this position.
On
Turbulent plumes, whether found in atmospheric gas dispersion or within high-energy fuel combustors, are central to applications across engineering, environmental science, and physics. These complex flows impact everything from air quality monitoring to fuel efficiency in combustion processes and are essential for fault diagnostics in aerospace propulsion. This project aims to explore whether it is feasible to image the concentration and temperature profile of gases within a turbulent plume using sparse optical measurements? Addressing this question requires modelling and analysis on how photons interact with highly dynamic media, combining principles from ill-posed inverse problems, and partial differential equations, such as radiative and heat transfer. Aside the harsh environment constraint that limits the amount of data that are feasible to collect, this endeavour is further complicated by the turbulent nature of the media itself. The challenge lies not only in reconstructing the complex features of these images from limited noisy data but also in establishing a rational way of parameterising and characterising the turbulence to suppress the uncertainty in the reconstructed images.
The research will delve into the study of random media and Bayesian inference frameworks as turbulent flows exhibit chaotic behaviour, and capturing data-consistent estimates of uncertainty will ensure the credibility of the reconstructed images. Moreover, the project offers an exciting opportunity to explore contemporary approaches in data-driven modelling of turbulence. Neural networks, including neural operators and diffusion models, show promise for efficiently representing the complex swirling and chaotic structures of turbulent flows, potentially offering an alternative or complement to traditional models.
This project provides a stimulating environment for mathematically inclined graduates, offering direct applicability to high-impact fields and the chance to contribute to pioneering research that bridges theory, computation, and practical application. Opportunities for direct collaboration with experts in gas metrology and aerospace to refine methods and validate results.
The University of Edinburgh is committed to equality of opportunity for all its staff and students, and promotes a culture of inclusivity: https://www.ed.ac.uk/equality-diversity
A first-class honors degree (or International equivalent) in Mathematics, Statistics or Computer Science, and preferably an MSc degree on a relevant topic, e.g., signal processing, data science.
An understanding of statistical inference is necessary. Experience with computational fluid dynamics/mechanics would be advantageous but not essential.
Further information on English language requirements for EU/Overseas applicants.
Applications are welcomed from self-funded students, or Home students who are applying for scholarships from the University of Edinburgh or elsewhere.
Off
Sensor networks, sensor fusion and management techniques address key challenges in intelligence, surveillance, target acquisition, and reconnaissance (ISTAR). Opportunities in adaptive data-driven sensor tasking and resource management include adaptive sensor placement, adaptive waveform design to reflect the target reflection characteristics and channel environments, and adaptive sensor selection. Although these problems have solutions in specific use cases, this theme will consider scenarios with broader applications involving multiple heterogeneous sensors on single or multiple cooperative autonomous airborne platforms.
The solutions developed in this should be robust to dynamic and congested environments, adverse weather conditions, and mutual sensor interference. A range of algorithmic and signal processing or machine learning technologies will be considered, as well as specific technical challenges. For example, projects in this theme will consider aspects related to
- wide area motion imaging (WAMI), position, navigation, and timing issues (PNT);
- robustness to adversarial attack;
- sensor fusion and tracking applications;
- use of kernel and Monte Carlo methods;
- outlier-robust (and other metrics) messages in belief propagation algorithms;
- and scheduling in large dynamic networks
Probabilistic and Bayesian frameworks will be preferred to enable uncertainty quantification and management.
The techniques, solutions, and challenges proposed in this theme has applications across a range of defence and civilian applications, including search and rescue, law enforcement, and remote sensing.
This project will be jointly supervised by:
- Prof James Hopgood, James.Hopgood@ed.ac.uk
- Prof Mike Davies, Mike.Davies@ed.ac.uk
Smart Products Made Smarter
The PhD project forms part of a larger Prosperity Partnership Programme, Smart Products Made Smarter, a collaboration with Heriot-Watt University, University of Edinburgh and Leonardo.
We are pleased to invite applications for a PhD studentship to work as part of a leading team of experts. This studentship will be supported by an enhanced stipend of £20,716 per year over 3.5 years.
This grant, sponsored by the EPSRC, is a collaboration between academia and Leonardo. There are currently PhD opportunities available to work on diverse topics as part of this collaborative team. The work will involve strong links with industry.
The research addresses a broad range of challenges. These challenges exemplify future product lifecycle management from smart concept, design, development and manufacture to enhanced end-user capability, united by a common digital thread to enable smarter products to be made smarter. Each challenge area has clearly identified initial research themes and associated research challenges to be addressed and these are indicated below:
Challenge 1 (C1) the Making challenge: To create new hybrid manufacturing processes, that combine multiple Additive Manufacturing (AM) process with precision machining and coating processes to create components that disrupt the traditional functional trade-offs of Size, Weight and Power (SWaP) through techniques such as varying the material properties within a part and harnessing the digital production of optical components.
Challenge 2 (C2) the Manipulation challenge: To create new handling processes that fully exploit the digital data flows which define custom components whose shape and functionality is tailored to production by dexterous, highly adaptable robots that are programmed dynamically.
Challenge 3 (C3) the Computation challenge: To create new signal processing & machine learning methodologies that enable intelligent, digital & connected sensor products while mitigating the data deluge from the multiple sensors produced by Leonardo operating across the EM spectrum.
The themes represent areas that could form the basis of your PhD. These PhD positions offer great flexibility and we welcome the opportunity to explore other ideas & themes
Please note that this advert will close as soon as a suitable candidate is found.
The University of Edinburgh is committed to equality of opportunity for all its staff and students, and promotes a culture of inclusivity: https://www.ed.ac.uk/equality-diversity
Please note that as this is a defence based project, only UK/EU students are eligible to apply. International applicants are not eligible.
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. Further information on English language requirements for EU/Overseas applicants.
Tuition fees + stipend are available for applicants who qualify as:
a UK applicant
an EU applicant (International/non EU students are not eligible)
Funding is available through EPSRC Prosperity Partnership Programme. As this is a defence related project there are nationality restrictions (see above).
Off
The Institute of Data, Imaging and Communications is seeking a PhD candidate to explore new approaches for numerical simulation using the probabilistic framework of Monte Carlo geometry processing. This research will focus on parabolic models, particularly the advection-diffusion equation that governs gas plume dispersion in congested urban settings.
Monte Carlo geometry processing methods offer a highly flexible, parallelisable, and mesh-free approach, making them appropriate for simulating PDEs in complex or dynamically changing geometries. Unlike traditional methods such as Finite Element or Finite Difference schemes that require domain discretisation, Monte Carlo approaches allow for modelling on the edge, bypassing the need to invert large and potentially ill-conditioned matrices. Originally developed for efficient rendering in computer vision, Monte Carlo methods are now being adapted to solve PDEs across a variety of fields. Leveraging recent advancements like the walk-on-spheres and walk-on-stars algorithms, designed for elliptic models, this project aims to deliver fast-converging variants of such methods for time-dependent parabolic problems.
A key challenge in this is to establish an analogue to the mean-value property in space-time domains. Moreover, whilst Monte Carlo methods offer significant advantages in accommodating dynamic boundary conditions, they also present unique challenges. Achieving fast convergence and managing statistical noise are ongoing areas of research, as these factors are crucial for applications requiring precision and computational efficiency.
This probabilistic framework also supports inverse problem-solving, such as detecting plume sources, which involves inferring release characteristics in geometrically complex environments. Synergistically, Monte Carlo methods allow for inherent uncertainty quantification, and this is particularly useful in situations with sparse measurements or stochastic model behaviours. The topic is also amenable for cross-fertilisation of ideas from randomised numerical linear algebra, exploiting the low-dimensional structure in the kernel of the advection-diffusion model.
Potential applications for this research include environmental monitoring and national security in urban settings. This project is well-suited for candidates with a strong foundation in mathematics, stochastic processes, and Monte Carlo methods.
The University of Edinburgh is committed to equality of opportunity for all its staff and students, and promotes a culture of inclusivity. Please see details here: https://www.ed.ac.uk/equality-diversity
A first-class honors degree in Mathematics, Statistics or Computer Science, and preferably an MSc in a closely related topic, e.g. data science or computational and applied mathematics.
A background in stochastic differential equations is necessary.
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. Further information on English language requirements for EU/Overseas applicants.
Applications are welcomed from self-funded students, or Home students who are applying for scholarships from the University of Edinburgh or elsewhere.
Off
Image
Soft robotics represents a transformative approach to creating robots that are inherently safer and more adaptive to real-world environments. Unlike traditional rigid robots, soft robots can deform, stretch, and interact with humans or delicate objects in a more natural and flexible manner. As these robots become increasingly vital in fields ranging from healthcare to industrial automation, one of the key challenges is developing advanced sensing technologies that seamlessly integrate with soft, flexible materials to provide real-time feedback for precise control and decision-making.
This PhD project aims to address this challenge by developing integrated fabrication techniques that embed sensing technologies directly into soft robotic structures. By combining expertise in flexible electronics, advanced materials, and 3D-printing-based fabrication methods, this project will enable the creation of multifunctional soft robotic systems with enhanced sensing capabilities. These systems will be capable of detecting touch, pressure, deformation and other physical parameters crucial for dynamic interaction with their environments.
This project will be carried out at the SMART Group within the Institute for Imaging, Data and Communications (IDCOM) and closely collaborate with the leading robotics researchers from the Soft Systems Group. This research provides an exciting opportunity to work at the intersection of materials science, robotics, electronics and machine learning. The ideal candidate will have a background in electronics/mechanical engineering, materials science, robotics, or a related field, with a strong interest in soft robotics and sensing technologies. By joining this project, you will contribute to the rapidly growing field of soft robotics and help develop systems that could reshape healthcare, manufacturing, and other industries.
Due to a multidisciplinary nature of this project, it will be supervised by an academic team that have expertise in different engineering areas. The supervision team will consist of:
Principal supervisor:
Dr Yunjie Yang (Imaging, Data and Communications), https://www.eng.ed.ac.uk/about/people/mr-yunjie-yang
Assistan supervisors:
Professor Adam Stokes (Integrated Micro and Nano Systems): https://www.eng.ed.ac.uk/about/people/professor-adam-stokes;
Dr Francesco Giorgio-Serchi (Integrated Micro and Nano Systems): https://www.eng.ed.ac.uk/about/people/dr-francesco-giorgio-serchi
Dr Michael Chen (Bioengineering) https://www.eng.ed.ac.uk/about/people/dr-michael-xianfeng-chen
Please also note that this advert will close as soon as a suitable candidate is found. Successful candidate will be expected to start in September 2025.
Please make sure to submit a research proposal as part of your application. For advice on writing a Research Proposal please see here:
https://www.ed.ac.uk/files/imports/fileManager/HowToWriteProposal090415.pdf
The University of Edinburgh is committed to equality of opportunity for all its staff and students, and promotes a culture of inclusivity. Please see details here: https://www.ed.ac.uk/equality-diversity
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.
Further information on English language requirements for EU/Overseas applicants.
Tuition fees + stipend are available for applicants who qualify as Home applicants* but exceptional international students will also be considered.
Applications are also welcomed from self-funded students, or students who are applying for scholarships from the University of Edinburgh or elsewhere.
Home Students
To qualify as a Home student, you must fulfill one of the following criteria:
• You are a UK student
• You are an EU student with settled/pre-settled status who also has 3 years residency in the UK/EEA/Gibraltar/Switzerland immediately before the start of your Programme.
Off