ESPRC Summer Research Placements Projects | Faculty of Science, Agriculture & Engineering

We have 34 projects listed below – grouped by School.

These are based in the Schools of Computing, Engineering, Maths, Stats & Physics, and Natural & Environmental Sciences.

If you are interested in finding out more about a specific project, please contact the named supervisor below.

Each placement is for 8 weeks. For further information on how to apply - please see the placements webpage.

If you have any questions, please contact doctoral.awards@ncl.ac.uk

School	Project code	Supervisor	Placement project title	Placement description
Computing	Comp01	Jaume Bacardit	Explainable AI for the identification of biomarkers of disease progression in osteoarthritis	This is a biomedical data science project focusing on a disease, osteoarthritis, for which there is currently no cure. We have access to a very rich longitudinal dataset on this disease that was created as part of the €15M IMI-APPROACH project, in which we have done some groundwork in designing AI models to predict disease progression (i.e. will the patients be worse off in two years' time?). The work of this placement involves the integration and harmonisation of this data to other available public datasets for this disease in order to validate our previous work, and the creation of Explainable AI interfaces to let non-technical audiences interact with our AI models
Computing	Comp02	Peter Bradshaw	An algebraic approach to single- and multi-qudit systems	Recent theoretical developments have made possible a purely abstract algebraic description of an arbitrary n-level quantum system in terms of meaningful physical observables in a manifestly symmetric way. This offers the potential for new interpretations of the states of such systems along more physical lines, independently of the traditional probability-based formalism of quantum theory. This method offers considerable discovery potential in the areas of quantum foundations and quantum information theory. The aim of this project is to study single- and multi-qudit systems along these algebraic lines, and to apply these insights to some simple quantum circuits. Time allowing, the study of more complex circuits or generalising the insights gained to systems with more levels is possible. This work will be within non-commutative algebra, and conducted along formal mathematical lines, potentially assisted by computer algebra systems. This project is well-suited to a student of computing or physics with strong mathematical and quantum inclinations, or to students of mathematics with an interest in fundamental physics.
Computing	Comp03	Wanqing Zhao	User-centred decision support tools for heat decarbonisation	This placement will support our ongoing work to develop a non-intrusive, data-driven approach for estimating the long-term costs of heat pump ownership. The role will interface with a team of researchers and an industry collaborator (Northern Gas Networks) from an EPSRC IAA project, thus participating our weekly project meetings. During the placement, the student will focus on the design and development of intuitive dashboards, aimed at helping prospective heat pump adopters explore cost projections, energy usage patterns, and the impact of behaviour changes. This will involve transforming complex AI model outputs and energy datasets into accessible and interactive visualisations that enable informed decision-making in the transition to low-carbon heating solutions. By end of the placement, the student will have gained hands-on experience in data-driven software development, user-centred design, and sustainable energy innovation.
Computing	Comp04	Ken Pierce	Feasibility study for structured acyclic nets (SANs) and machine learning	Systems based on machine learning (ML) are becoming ubiquitous. Many of these systems are 'black boxes' to their end users, because currently there is no logical framework for reasoning about their decision-making processes. The absence of such a framework undermines confidence in ML-based systems and in their decisions. Causal reasoning is a key feature of science, engineering, and rational investigation in general. We are developing a proposal for ESPRC on using structured acyclic nets (SANs), a visual formal method based on Petri nets, to model causal relationships in ML projects. We aim to relate human and ML decision-making process models using the behavioural and temporal abstraction mappings of SANs. The overall goal is to demonstrate whether a SAN could encode a simple machine learning algorithm, such as the classic four-neuron XOR problem, and assess the potential merit of the approach. The SAN would be encoded using the Workcraft tool, developed at Newcastle under several EPSRC projects. Depending on progress, there are several forms of net-based machine learning that could be explored, including neutral networks and Tsetlin machines. It may also be possible to explore extending Workcraft, depending on the selected intern's interest and skillset.. Specific steps in the internship would be: - Onboarding and arrangements for working pattern, location, and meetings - Reading about SANs and machine learning algorithms (depending on existing skills and knowledge) - Following the Workcraft and/or machine learning library tutorials (depending on existing skills and knowledge) - Manually building a SAN for a basic four-neuron neural net for the XOR problem
Computing	Comp05	Varun Ojha	Vision and language model based eyeware for low vision people	Assistive Technology is making people's lives better. Open source glasses are helping blind or low vision persons. This project uses an open-source AI wearable platform designed to function as a "second brain" for enhanced productivity and memory. The core system involves an Omi hardware device (goggles) that continuously records audio, which is then streamed to a mobile/desktop application and a powerful Python/FastAPI backend. This backend uses various AI services to automatically transcribe conversations in real time and generate structured data, including summaries, key insights, and action items. The entire ecosystem is built to be open and extensible, allowing developers to contribute to the code and create a wide array of plugins and integrations for connecting Omi's captured data with other popular services like Notion, Slack, and ClickUp. The main objective of this project is to enhance this open-source technology for various purposes.
Computing	Comp06	Deepayan Bhowmik	AI Foundation model for environmental monitoring in lough neagh	Lough Neagh, the United Kingdom's largest freshwater lake, has experienced severe invasion of the harmful algal blooms (HABs) that compromise drinking water quality and ecosystem health. This student placement aims to support an ongoing long-term project on the development of a new type of artificial intelligence system by building a foundation model for monitoring of algal bloom using remote sensing and AI. The student will assist the team members with data preparation and algorithmic testing of a new multi-modal transformer with a hybrid physics-informed AI model that will harmonise and unify all environmental information from Earth observations, environmental measurements, and ecological surveys. The algorithm will serve as a general-purpose base model on which multiple environmental data sets can be modelled, leading to greater understanding of our ecosystems and climates of Lough Neagh. During the placement, the placement student will learn through exposure to the development of the state-of-the-art AI models to address one of the key environmental issues in the UK. The student is expected to carry out the following tasks: 1. (Week 1-2) Literature review of the foundation AI models and environmental monitoring of Lough Neagh 2. (Week 2) Familiarisation of the tools and techniques for remote sensing data processing (such as satellite imaging). 3. (Week 3-6) Data gathering from available sources and ground truth preparation of the AI model development. Development of some code to assist the main algorithm development. 4. (Week 6-7) Testing the algorithms (developed by key researchers/PhD students) and performing qualitative and quantitative analysis and 6. (Week 8) Write a report of about 2,000 words.
Engineering	Eng01	Ekin Erkus	Inside the forecast: explaining predictions with multi-agentic AI	This placement develops a small research prototype that helps make machine learning forecasts easier to understand. Many forecasting models can produce accurate predictions, but it is often difficult to explain clearly which parts of the input data influence a prediction and how the model reaches its final output. This project addresses that problem by building a multi-agent large language model framework that converts internal model evidence into clear written explanations for both technical and nontechnical users. The project uses a Tsetlin Machine, which is a rule-based machine learning method that makes predictions by combining many simple logical patterns learned from data. Although this structure offers good potential for explainability, the raw information produced during prediction is still difficult to interpret directly. The placement therefore focuses on turning this internal information into accessible explanations. The work is organized around three linked agents. First, a Model Evidence Agent records and summarizes which learned rules are most active for a given forecast and which input features contribute most strongly to those rules. Second, a Context Analysis Agent compares these internal signals with patterns in the dataset and with supporting notes prepared during the placement, so that the explanation can be related to observable trends in the data. Third, an Output Agent converts the collected evidence into two forms of explanation: a short technical summary and a plain language summary for nontechnical users. The project is intentionally scoped for eight weeks and is designed to be achievable for an undergraduate student. Training large models from scratch is not required. Instead, the placement focuses on structured data logging, agent design, evidence organization, prompt design, and small-scale evaluation on a bounded forecasting case study. By the end of the placement, the expected outputs are a working prototype, a small evaluation set with example explanations, and a short written report.
Engineering	Eng02	Manuel Herrera	A global benchmark database of urban street networks: structural comparison across continents	Understanding how street networks differ across the world is fundamental to urban planning, transport modelling, and infrastructure resilience research. Differences arise across continents, cultures, climates, and city sizes. However, no standardised and openly available benchmark database currently exists that systematically compares street network structure for small and medium-sized cities across continents. This project will provide the student with hands-on training in computational urban analysis, a first view of reproducible research workflows, and the benefits of using open (urban) data. This project will use OSMnx, a well-established Python library for retrieving and analysing street networks from OpenStreetMap. The student will select a representative sample of small and medium-sized cities (population approximately 20,000-100,000) across Africa, America, Asia, Europe, and Oceania. Their street networks will then be extracted using a consistent and reproducible workflow. To support the student in getting started, the supervisor will provide an initial Python template script. This script will demonstrate the core workflow for retrieving a street network using OSMnx, computing a set of basic structural indicators, and exporting the results into a structured dataset. The student will then extend this template to analyse a larger set of cities, ensuring consistency and reproducibility across the dataset. The project will also include a non-computational component through structured online engagement. The student will hold short meetings with urban researchers and planning practitioners from different regions, facilitated through the supervisor’s academic network. These discussions will provide feedback and reassurance to the student, helping to contextualise the computational findings and to validate the selection of cities. The student will also hold weekly meetings with the supervisor. The student will deliver an online presentation of results to stakeholders, supervisor, and other interested parties.
Engineering	Eng03	Andrew Aspden	Entrainment in self-similar unsteady turbulent jets	Entrainment in self-similar unsteady turbulent jets is relevant to a wide range of environmental, biological and technical systems; for example, fuel injection in internal combustion engines, where inherent unsteadiness affects jet penetration, entrainment and mixing, with consequences for ignition timing. Physical prototypes are expensive, and numerical simulations can be prohibitively computationally demanding. Mathematical analysis of entrainment in turbulent jets can be used to predict mixing behaviour and develop low-cost modelling approaches that can be used to explore a broad parameter space rapidly. In general, the models lead to unique well-defined solutions, but a particular equation set (based on so-called mass-weighted species age transport) suffers from a curious issue of admitting any power law solution. This project aims to resolve this issue by exploring the consequences of modelling assumptions (for example, self-similarity, entrainment coefficient, and turbulent diffusion modelling) on the resulting solutions to the equations of motion. The resulting reduced models will require implementation (programming) of a bespoke numerical solution. Support with programming and high-performance computing will be offered, along with interaction with group members (PDRA/PGR), and involvement in group meetings.
Engineering	Eng04	Matthew Dyson	Programming a real prosthetic hand	The student will be working in a lab environment, programming a prosthetic hand with the team who developed the electronics and firmware. Work will take place in the Intelligent Sensing Lab, and the student will have the opportunity to attend lab/team meetings across multiple biomedical engineering labs in the School of Engineering. To facilitate networking, and to obtain a better understanding of the field, there will be an opportunity to visit a start-up developing a new upper-limb prosthesis. Currently, the hand has basic grip functionality. The most pressing requirement for enhancement is some form of interface, whether wired, Wifi or Bluetooth. The student will decide what functionality to add and how they want to do it, explain the reasoning to the wider team, implement the new firmware, and if time permits assess the resulting system. All the hardware for this project is already in place. All development work will be done using open-source software. The project is appropriate for a student with middle to advanced programming skills in C or C++.
Engineering	Eng05	Vladimir Zivkovic	Beyond Geldart: a new particle classification for intensified toroidal fluidization	Fluidization is a pillar of particle technology, providing the high heat and mass transfer rates essential for modern chemical and mineral processing. While the traditional Geldart Classification is the standard for conventional beds, it proves insufficient for Process Intensification (PI) systems like the Torbed reactor. By employing a high-energy swirling gas stream, the Torbed reactor dramatically intensifies heat and mass transfer, allowing for significantly faster processing within a much smaller footprint. Current industrial knowledge is largely centred on compact beds of coarse, millimetre-sized particles—primarily used for high-efficiency drying—where the swirling motion is directly tied to the minimum fluidization velocity. However, our preliminary research shows that for fine particles, such as 70 um silica, the physics shifts entirely. Rather than a stable, compact layer, these finer particles create an expanded bed that behaves more like fast fluidization, where the minimum swirling velocity is governed by the particles' terminal velocity rather than their minimum fluidization velocity. This research project aims to map this "hidden" transition point by analysing a range of particles from 50 um to 700 um. Establishing this new classification is vital; while compact beds excel at drying and solid processing in general, expanded beds hold massive, untapped potential for advanced gas processing, including carbon capture and emissions control. This study will provide the fundamental data needed to design energy-efficient, scalable reactors capable of handling the next generation of fine-particle industrial applications.
Engineering	Eng06	Mark Geoghegan	Quartz crystal microbalance investigation of surface coatings	We have developed an adhesive formulation that can be made into thin conductive films. These films will swell a little in water and we are interested in measuring this to identify how much liquid is retained by the films. The Quartz Crystal Microbalance (QCM) is a sensitive technique that can determine mass changes in films of the order of several tens of nanometres or less. It is possible that the project would require a small amount of chemistry, which will be shown to the student. The basic formulations will have been prepared, but some further modification to make them conductive will be needed. The QCM is new and it requires commissioning. The student will therefore get hands-on experience of a technique that has not been used for any previous project.
Engineering	Eng07	Jennifer Olsen	Evaluating wrist based wearable sensors for detecting orthostatic heart rate responses	This project aims to use low cost sensors that measure blood flow and motion to assess whether clinically useful data for the diagnoses of medical conditions can be acquired. These findings would form proof of concept data for a larger study exploring whether similar technology could provide clinically acceptable blood pressure estimates to help diagnose (or exclude) Postural Orthostatic Tachycardia Syndrome (PoTS), which is a type of dysautomnia and affects around 100,000 people in the UK. The student will build and test a simple wrist based sensor to measure how the body responds when a person moves from lying down to standing up. Using a small optical device that shines light into the skin, they will record heart rate and blood flow signals alongside movement data. Volunteers will be recruited to gather data, where the participants will move from a seated to standing position while the sensor collects information. The student will then analyse these signals to see whether clear, repeatable blood-flow changes can be detected during the posture shift. This small project will show whether affordable wearable technology could help monitor conditions that affect blood flow and heart rate. Therefore, the project will require the student: • uses commercially available hardware to build a sensor system capable of assessing heart rate and movement • collects in-person data from (healthy, able-bodied) volunteers (ethics approval 73557/2026) • writes code to analyse the data and validate the sensor as a diagnostic tool Ideal skillset: • coding experience, preferably C++ • interest in biomedical/health tech
Engineering	Eng08	Tousif Rahman	Automated hardware acceleration of embedded signal processing using logic gate networks on low-power FPGAs	This project investigates the automated generation of custom hardware accelerators from domain-specific code, targeting resource-constrained FPGAs such as those produced by Lattice or Effinix. The goal is to move away from manual hardware design workflows, which require significant expertise and engineering effort, towards a system that can analyse standard C/C++ code, identify computational kernels suitable for acceleration, and synthesise corresponding hardware implementations automatically using Logic Gate Networks (LGNs). To the best of our knowledge, this represents the first application of LGNs specifically to edge accelerator design, and the proposed toolchain opens up an entirely new use case for the emerging class of ultra low power FPGAs that have until now lacked the automated design flows needed to unlock their full potential. Students will gain hands on experience across both software and hardware engineering. This is a rare opportunity that most undergraduate programmes do not provide. The project spans the full stack from code analysis through to FPGA implementation and validation. On the technical side, students will learn how to use FPGA design tools and open source synthesis toolchains. They will build and train Logic Gate Networks and understand how they work. They will learn how software algorithms can be mapped directly onto custom hardware. They will also develop practical signal processing skills through the EMG simulation work and will strengthen their C/C++ and Python programming in a real research setting. Students will work independently on an open ended research problem. They will make real design decisions and present their reasoning to an experienced team. Regular meetings and progress presentations will build confidence in communicating technical ideas clearly. The engagement with PragmatIC will give students a direct view into the semiconductor industry. They will see how academic research connects to real commercial products. This is particularly valuable for anyone considering a career in hardware engineering, embedded systems or deep tech.
Engineering	Eng09	Jingjing Zhang	Agentic AI for automatic biomarker discovery from microsampling data	The student will design and build an adaptive AI agent that automates biomarker discovery from microsampling datasets. The agent will detect dataset structure, choose preprocessing steps, run simple/ML/Statistical models, compare multiple explainability methods, and generate structured outputs. The student will also develop an interactive dashboard to explore biomarkers, compare groups, and visualise explanations. This placement will be embedded within the EU-funded COMFORT project, which aims to build more accessible and personalised approaches to testing and monitoring illness using Patient Centric microSampling. To gain an understanding of the wider research context, the student will: • attend weekly Newcastle-based COMFORT project meetings to understand the broader microsampling and AI landscape • gather informal feedback from researchers and PhD students to shape the agent and dashboard • participate in networking opportunities with COMFORT partners (e.g. GSK, Pfizer, AstraZeneca, Lilly, Collaborate Health Cloud, Karolinska Institute). The project is structured to encourage autonomy, creativity, and informed judgment. The student will take the lead in selecting appropriate models, choosing and evaluating explainability techniques, designing the dashboard interface, and determine how the agent should adapt to new datasets. Students with additional interest may optionally explore agentic-AI frameworks (e.g. LangChain or similar open-source tools) to extend the agent’s reasoning or explanation capabilities.
Engineering	Eng10	Oktay Cetinkaya	Sensing the silence: detecting and diagnosing anomalies in IoT water meter networks	Every day, UK water networks lose over three billion litres of treated water due to leaks that are difficult to detect and localise. While IoT-enabled smart meters are increasingly used to monitor water usage and hence the leaks, the data they generate is often unreliable in practice. Sensors may fall silent or behave intermittently due to environmental conditions, physical obstructions, or network issues, making it difficult to distinguish between communication failures and genuine infrastructure faults. This project explores how such anomalies can be detected and interpreted using real-world datasets provided by Connexin/Northumbrian Water. The student will analyse spatial, temporal, and contextual patterns (e.g. weather and local activity), develop data-driven and signal processing-based methods (with optional machine learning components) to identify abnormal reporting behaviour, and investigate how anomalies relate to real-world conditions. Typical activities include analysing historical datasets, identifying patterns of normal and abnormal behaviour, and exploring relationships between neighbouring sensors and environmental factors. The project includes engagement with industry partners and, where feasible, field-based insight into real deployments. It offers a hands-on introduction to data-driven sensing, smart infrastructure, and practical challenges in large-scale IoT systems, with scope for the student to contribute their own ideas. Essential skills: basic programming (e.g. Python), interest in data analysis, and ability to work independently. Desirable skills: familiarity with data handling or visualisation tools, and interest in machine learning or IoT systems.
Engineering	Eng11	Martin Johnston	Design of quantum error-correcting codes	This project will investigate different methodologies to construct new quantum error-correcting codes suitable for use in future quantum communications and quantum computing. This work strongly aligns with the objectives of the University's Quantum Strategy Group and brings together the expertise of two schools, Engineering and MSP. It will contribute greatly to the growth of quantum computing in the university which is an important global research area. The transmission of quantum bits (qubits) is significantly affected by the presence of severe quantum noise which causes the state of the qubit to change and the information is lost. Quantum computing is not possible unless information-carrying qubits are protected with error correction techniques. This is a simulation-based project using free quantum computing software such as Pennylane or IBM's Qiskit. We will introduce the student to the fundamentals of quantum error correction and train them to use the software. The student will meet each week with the supervision team to discuss progress and actions. We will work together to look at existing cutting-edge classical error-correcting codes with the aim of modifying their encoding and decoding algorithms to operate on qubits affected by quantum noise. There is great scope for novelty in this project and it is very likely to results in at least a conference publication. We treat the qubits as ideal abstract mathematical objects with certain properties so experience in quantum mechanics is not necessary, although it would of course be advantageous.
Engineering	Eng12	Xiang Xie	Hardening AI-generated policy evidence: statistically robust adjudication of an urban knowledge graph	The objective of this placement is to apply a rigorous three-stage validation protocol to a large-scale, AI-constructed Knowledge Graph (KG) for urban environmental governance. The project aims to transition the KG from a preliminary prototype to a production-ready evidence base by (1) quantifying expert agreement using inter-rater reliability statistics, (2) adjudicating complex socio-technical relationships through direct expert engagement, and (3) enriching the graph with validated socio-demographic and enforcement cost data. To provide an understanding of the wider context beyond pure modelling, the student will conduct structured interviews and surveys with practitioners. These interactions will allow the student to validate the real-world 'mechanistic plausibility' of the Knowledge Graph's claims against professional expertise. The student will function as a Knowledge Engineer and Social Researcher. Key tasks include: • methodological adjudication: applying a structured decision rubric to "salvage" thousands of relational triples by reformulating abstract entities or decomposing compressed causal chains into explicit pathways. • expert engagement (contextual field work): The student will move beyond pure modelling by designing and conducting structured interviews and questionnaires with domain experts (e.g. urban planners, local government air quality officers, and environmental scientists). These interactions will serve to validate the "mechanistic plausibility" and "contextual dependence" of specific relationship claims within the KG. • statistical analysis: using Python or R to calculate Fleiss’ Kappa and Cohen’s Kappa to quantify the level of agreement between the AI’s output and human expert judgment, identifying systematic patterns of machine-human misalignment. • knowledge integration: translating qualitative insights from the fieldwork into new nodes and properties in a Neo4j graph database, ensuring the KG reflects the practical constraints of policy implementation.
Engineering	Eng13	Shengyu Duan	Designing ultra-low-cost RISC-V processor with custom instruction extensions for logic-based machine learning acceleration	This project aims to design a general-purpose RISC-V processor with custom instruction extensions to accelerate logic-based ML algorithms at lower cost than existing commercial solutions. The expected outcome includes a complete RTL design of the processor core, suitable for implementation on both FPGA and ASIC platforms, as well as a post-layout design (GDSII format), with the potential for fabrication using flexible IC technologies and eventual tape-out. The specific tasks include: • instruction analysis and compilation: Analyse typical software implementations of logic-based ML and develop application-specific instruction sets with custom extensions. • RTL design: Design and implement the RISC-V processor with logic-based ML acceleration, building upon the open-source SERV processor, a bit-serial RISC-V core optimised for minimal area. • physical design flow: Execute the full IC design flow, including synthesis, placement and routing, to generate a post-layout implementation. • evaluation: Assess performance, power, and area, and compare against existing RISC-V-based designs. Through this project, the student will gain hands-on experience in: • firmware development and deployment for embedded ML applications • RISC-V processor design and toolchain usage, including compilation, customisation, and instruction extension support • end-to-end IC design flow, from RTL design to synthesis and place-and-route
Engineering	Eng14	Narakorn Srinil	Data-informed fluid-structure interaction simulation using phenomenological model	This student placement project focuses on data-informed fluid–structure interaction (FSI) simulation for predicting vortex-induced vibration (VIV) in engineering systems. VIV occurs when fluid flow past a structure generates alternating vortices that induce oscillatory forces on the structure, potentially leading to fatigue damage in offshore cylindrical structures such as subsea dynamic cables and monopile foundations of wind turbines exposed to currents. Reliable prediction of VIV response is therefore essential for the safe and efficient design of fluid–structure systems. The student will build a structured database of VIV experimental results and perform data preprocessing, feature engineering, and exploratory data analysis to understand relationships between fluid–structure parameters and observed vibration responses. Machine learning techniques may be used to assist in identifying patterns in the data, estimating empirical coefficients, and improving the calibration of the phenomenological model. The project will involve implementing computational tools in MATLAB to perform model training, parameter optimisation and validation. Techniques such as regression models, parameter fitting, cross-validation, and performance evaluation metrics will be applied to ensure the robustness and generalisation capability of the calibrated model. By the end of the placement, the student will gain practical experience in scientific data collection, machine learning workflows, model calibration, and fluid–structure interaction analysis. The project will provide valuable exposure to data-driven modelling approaches in engineering and may contribute to improved predictive capability of phenomenological VIV models used in FSI research.
Engineering	Eng15	Matthew Deakin	Enabling increased connections of household PV and electric vehicles through flexible power converters	The number of electrical vehicles and solar PV systems being connected by households is resulting in a fast increase in the power capacity requirements of substations in electrical power systems. As a result, supply chain shortages have rapidly increased the costs of critical components in substations, particularly transformers. At the same time, public opposition to new energy infrastructure (including substations) is growing. To-date, a prototype of the approach has been designed and built, but only the most basic operation has been demonstrated. Working alongside researchers in Newcastle’s state-of-the-art Smart Grid Lab, this project will connect the flexible power converter to a new transformer test rig and demonstrate the full benefits of the novel flexible power converter. Ultimately, each such device could enable the connection of 20 domestic electric vehicle chargers. Candidate requirements: We are looking for an enthusiastic student who is interested in engaging with a lab-based research project in either energy, power engineering, or electronics. The ideal student would be studying engineering or a numerate subject (e.g., physics, computing). Expected project schedule* Weeks 1-3: Developing simulations of the flexible power converter. Inductions to Smart Grid Lab. Introduction to the flexible power converter control implemented in the lab to-date; working to transfer operation to the new test rig. Weeks 7-9: Developing controls showing full operation capabilities. Week 8: Finalise project and summarise results. *There can be some flexibility to account for students’ own interests and/or holiday requirements - to be discussed prior to start.
Mathematics Statistics & Physics	MSP01	William Rushworth	Unknotting number is not additive	A mathematical knot is a circle embedded in 3-dimensional space. It turns out that all of the knotting can be removed by allowing the knot to pass through itself; the minimum number of times required to do this is known as the unknotting number of a knot. Two knots can be added together by cutting them open and splicing the ends together. This operation is known as connected sum. For over 100 years it was unknown if the unknotting number behaved nicely with respect to connected sum i.e. if the unknotting number of the connected sum is the sum of the unknotting numbers. In the summer of 2025 Brittenham and Hermiller proved that, surprisingly, this is false in general: there exist pairs of knots such that the unknotting number of their connected sum is strictly less than sum of their unknotting numbers. This project will further investigate the result obtained by Brittenham and Hermiller. Specifically, their result raises more questions than it answers: they simply found a pair of knots on which the unknotting number is not additive. There is no deeper understanding of why this result holds, or indeed what, if anything, is special about the particular knots they found (especially interesting given that the resisted discovery for a century). This project will search for such a deeper understanding, by attempting to find new examples of knots on which the unknotting number is not additive, and linking their existence to other topological objects. As is common in low-dimensional topology, this project is exceptionally well-suited for undergraduate students: the material is concrete and accessible, but links directly to many complex and far-reaching areas of mathematics. The student shall begin by assimilating the important background material, before understanding Brittenham and Hermiller’s result. They will then move on to exploring how this result can be generalised and extended.
Mathematics Statistics & Physics	MSP02	Leo Tsui	Optimising colloidal quantum dots for next generation photonic applications	Colloidal quantum dots are nanoparticles that exhibit unique optical properties due to quantum confinement effect. A wide range of optoelectronic applications, such as solar cells, LEDs, lasers and bio imaging have been developed. Their applications in quantum photonics, for example single photon source and quantum information processing, are being actively explored. Therefore, it is of interest to establish the synthesis-structure-property of these nanomaterials and tailor them for next generation photonic applications. In this project, we will investigate the influence of synthesis condition and composition on the optical properties of carbon-based colloidal quantum dots. We will develop an environmental-friendly synthesis method for carbon-based quantum dots and perform optical characterisation using spectroscopic techniques. The student will gain hands-on experience in material synthesis, optical measurement, programming and data analysis.
Mathematics Statistics & Physics	MSP03	David Kimsey	General measure theory and the Lebesgues-Stieltjes integral	Building on from the content of MAS3716 (Measure Theory), we will explore a general class of measures which contains the Lebesgue measure. We will go generalise various results on measures to this setting and reconstruct the Lebesgue integral with respect to an arbitrary measure in the class (this is called the Lebesgue-Stieltjes integral). We will then generalise various results on integration theory to the Lebesgue-Stieltjes setting.
Mathematics Statistics & Physics	MSP04	Aleksey Kozikov	Engineering and stabilising ultra-thin perovskite devices for quantum applications	This project will investigate the fabrication, stabilisation and electrical–optical characterisation of ultra-thin two-dimensional perovskite materials as candidate platforms for future quantum technologies. The primary objective is to experimentally evaluate how encapsulation strategies and device structuring influence the optical emission properties and electrical transport behaviour of exfoliated 2D perovskite flakes. The student will fabricate micron-scale devices using mechanical exfoliation and deterministic flake transfer techniques, followed by structural, optical and electrical characterisation. A central research question will be: How do encapsulation approaches (e.g. hBN protection layers and environmentally benign coatings) modify stability, emission efficiency and charge transport in ultra-thin perovskites? The student will analyse experimental data, compare results with current literature and propose optimisation strategies for improved device performance. Beyond laboratory work, the student will participate in weekly group meetings, engage in scientific discussion and receive training in research integrity and safe laboratory practice. The project is designed to provide scope for independent thought, including experimental design decisions, interpretation of data and refinement of fabrication strategies. The outcomes anticipated include a stability-performance map for encapsulated 2D perovskite devices and recommendations for scalable fabrication routes relevant to future quantum photonic platforms.
Natural & Environmental Sciences	SNES01	Andrew Benniston	A green approach to esterification: photocatalytic ester formation	Polyesters are an important class of polymers, produced by the esterification of an alcohol with a carboxylic acid in the presence of a suitable catalyst. The choice of catalyst is crucial since low temperature reactions are preferred, especially in industrial applications, for both practical energy sustainability and commercial reasons. For example, the production of polyester is performed on the 63 m tonnes scale every year. Considering the importance of the esterification reaction in countless commercial applications, new methods to promote the reaction in high yield are actively investigated. The objective of the project is to test a new esterification process based on LED light activation of a newly designed photocatalyst, representing a disparate method of carboxylic acid activation. The process would fit within the remit of green chemistry and also have potential commercial applications. The student would prepare the photocatalyst and test it out, with the scope to design their own experimental procedures and learn new spectroscopic techniques.
Natural & Environmental Sciences	SNES02	Tom Smith	Peptide & prejudice: revisiting ‘unsuitable’ chemistry for use in next generation peptide drug discovery	Background: peptides consist of long chains of amino acid building blocks and have been revolutionary in biological research, cosmetics, and drug discovery. New peptides for drug discovery can be cheaply, efficiently, and rapidly identified by screening vast (>10^12) peptide ‘libraries’ which are barcoded with an ‘RNA’ tag. This tag is large and notoriously unstable. For this reason, well established chemistry (e.g. Heck) has been avoided in peptide drug screening which severely limits its applications. However, a recent publication, and preliminary data generated by our group has shown that the RNA tag is more stable than previously thought. Aims & Methods: this short chemical biology placement will explore new applications of some common chemical reactions for peptide modification in the context of RNA-tagged screening libraries. The student will perform peptide synthesis using unnatural synthetic building blocks on industry-grade automated instruments. They will then use state of the art conjugation chemistry to attach the RNA barcode. Finally, they will choose and perform peptide modification chemical reactions and analyse the integrity of the peptide-RNA conjugate. Reactions which best preserve RNA integrity can be used in the future for the generation of modified peptide libraries for the discovery of next generation therapeutics. Skills: the student will gain interdisciplinary technical skills in small molecule, peptide, and bioconjugation chemistry, as well as biochemical techniques in RNA degradation assays. There is also significant opportunity for the student to improve their research and transferrable skills such as problem-solving/independence in determining chemical reaction choice, and communication within our highly interdisciplinary Newcastle chemical biology team with members from a variety of scientific disciplines. The student will also receive all necessary training and participate heavily in team meetings, being required to give a formal project presentation at the end of the placement.
Natural & Environmental Sciences	SNES03	Sam Wilson	Novel sensors for detecting harmful algal blooms	The placement student will be ground-truthing UK Space Agency-funded aquatic sensors for detecting algal blooms. The sensors will be deployed in the Lake District which increasingly suffer from algal blooms in the summer months. The work will involve (a) deployment and recovery of sensors in Lake Windermere (b) cross-comparison with wet chemical analysis (c) quality assessment of the datasets (e.g. variability, sensitivity, drift, biofouling).
Natural & Environmental Sciences	SNES04	Harriet Annabella Stanway-Gordon	Development of DNA-compatible Tsuji-Trost reactions for application in encoded library synthesis of structurally diverse macrocycles	This project aims to develop initial conditions for an intermolecular DNA-compatible Tsuji-Trost reaction and to undertake preliminary optimization of the system. Working in collaboration with the Waring group, this will build on their previous successes using micellar catalysis for DNA-compatible reactions, applying a similar approach in this instance. It is anticipated that preliminary results from this project will be used for future funding applications related to incorporation of the reaction into macrocyclic structures for DELs and generation of the subsequent libraries. The student will work on model DNA-conjugates, synthesising appropriate starting materials for use in subsequent reaction optimisation. They will then explore various reaction parameters to allow for DNA-compatible Tsuji-Trost realisation, monitoring and analysing outcomes via LC-MS. If deemed appropriate from preliminary investigations, they will also have the opportunity to design and perform factorial experiments. The student will gain hands-on experience of conducting an independent research project at the interface of chemistry and molecular biology, developing practical skills in DNA-handling and analysis alongside synthetic chemistry, reaction development and analytical techniques. The student will also benefit from one-to-one mentorship and supervision under an early career research fellow, allowing for insight and exposure to academic career pathways, within the wider framework of the busy and supportive Medicinal Chemistry Group (ca. 35 members).
Natural & Environmental Sciences	SNES05	Connor Gallagher	Optimising polysaccharide blends as biodegradable thickening agents	This project is part of a larger effort that aims to develop biodegradable—and ideally, naturally derived—alternatives to the polymers used for various functions in commercial beauty products. The beauty industry uses myriad water-soluble polymers to control product structure, stability, and flow behaviour during use. These polymers offer excellent tunability and greatly enhance the consumer experience; however, after use they accumulate in the environment and are difficult to degrade causing environmental concerns. This work aims to develop new biodegradable polymers that provide the same consumer benefits while minimizing end-of-life concerns. Specifically, the student will explore using blends of xanthan gum with other polysaccharides (e.g. guar gum, alginate, etc.) as alternatives to currently used synthetic polymer-based thickening agents. This project is ideal for an undergraduate student who is interested in pursuing a research career in either an industrial or academic setting. Scientific research is a different experience to laboratory work undertaken as coursework and this project will enable the student to gain critical experience in iterative thinking, problem solving, and answering open-ended questions that will be key in a future research career. This research is poised at the intersection of microbiology, polymer chemistry, and materials science and the student will gain experience across all three disciplines. In addition to “hard” laboratory skills, the student will also develop “soft” skills in teamwork, professionalism, and scientific communication. The commercial nature of the problem we’re solving will allow the student to experience first-hand how innovation in the laboratory can be translated into real-world impact.
Natural & Environmental Sciences	SNES06	Michael Carroll	Making the switch to ‘green’ solvents: Improving the safety and sustainability of the chemistry teaching laboratories	Project Summary: A recent review of the chemistry teaching laboratory (Stages 1–4 and PGT) identified multiple opportunities to improve both safety and sustainability in our practical classes. In particular, the annual use of dichloromethane (DCM, ~125 L/year) and diethyl ether (~145 L/year) present environmental, regulatory, and safety challenges. This project will evaluate and implementing sustainable alternatives for selected undergraduate experiments. Project activities: Over the 8-week placement, the students will: • identify a prioritised set of experiments that use DCM and/or ether • investigate and test safer, more sustainable alternatives (e.g., ethyl acetate, ethanol) for reaction, work up, and purification steps • collect experimental data comparing performance, safety, and environmental impact • work closely with academic staff, PGR demonstrators, and the technical team to refine procedures • update the relevant sections of the laboratory manuals in collaboration with the lab organiser • prepare a short report summarising the sustainability benefits and expected reductions in use of DCM and ether. • produce posters highlighting the “greening” of the practical classes
Natural & Environmental Sciences	SNES07	Lee Higham	Novel fluorescent phosphonium salts as multimodality probes for biological cell imaging.	One important application of fluorescence is in medical imaging; a fluorescent molecule is synthesised and taken up by a biological cell – fluorescence microscopy is then used to visualise and localise the compound. If this molecule is also cell or organelle specific, the research scientist can use the sensitivity of the technique to image these locations and determine whether they are functioning properly. Recently we have been developing metal-free multimodality imaging agents. The latest work in the LJH group has seen the development of a family of fluorescent bodipy dyes based on phosphonium salts. Treatment of a phosphine with a dihaloalkane causes an alkylation reaction which forms a new phosphorus carbon bond, rendering it cationic. The positive charge of the phosphonium salt causes these probes to accumulate in the mitochondria (which has a negatively charged membrane) and the bodipy core allows for its fluorescent imaging, which is the ‘battery’ of eukaryotic cells. If the mitochondrial function is abnormal, its charge is disrupted which in turn affects the incorporation of the sensitive probe. The student will prepare the compounds described above, gaining experience in chemical synthesis, purification and analysis. This will require the acquisition and development of technical skills including column chromatography and inert atmosphere Schlenk-line manipulations. All of the equipment required to accomplish these goals will be made available to the student, and training on the specialist NMR, IR and X-ray crystallographic facilities and UV-Vis/Fluorescence analytical instrumentation will be provided. Many of the consumables necessary for the successful completion of the work will be available as a result of my previous EPSRC funding, with the exception of those chemicals which will be consumed by the student during this project alone.
Natural & Environmental Sciences	SNES08	Tom Atkins	Understanding the impact of grassland management on wildflowers, insects, and ecological complexity through the use of network science	This exciting placement involves joining a team to undertake a range of biological fieldwork techniques rooted in network science. These techniques are widely used in industry and academia and are very useful skills to learn. This important project forms part of a larger initiative in collaboration with Northumberland County Council to improve urban amenity grasslands for both people and wildlife, minimising the trade-offs, and has the potential to contribute to informing management decisions across the entire county. In Northumberland, and across the UK, urban amenity grassland holds particular social importance through recreational, aesthetic, and historic values, and is typically associated with intensive management, often to meet the needs of local communities e.g.: for sporting areas. Yet intensive mowing regimes in urban grasslands have been associated with reduced wildflower and pollinator abundance and diversity, and arthropod abundance more generally. However, the picture remains unclear, with different groups of organisms The research placement would involve around eight weeks of partially-flexible ecological fieldwork within a fully replicated experiment. There are six study sites in easily-accessible towns in Northumberland. Students will learn how to undertake a range of ecological techniques, such as ground flora sampling, ground and foliar invertebrate sampling, soil sampling, Biodiversity Net Gain assessments (for those interested in environmental policy), and plant-pollinator transects. Students will also learn how ecological data can be used to construct networks in order to understand the complexity and resilience of the ecosystem.