STEM Summer Research - Strathclyde Courses

You will earn 6 research credits over 8 weeks, conducting faculty-supervised, hands-on, directed study research projects with results that will culminate in the preparation of a research paper.  You will complete a minimum of 240 hours on research in and out of the laboratory.

Faculty mentors will work closely with you to direct your continued growth and knowledge development in the chosen research topic discipline.

  • Please review your project with your academic or study abroad advisor to ensure it will transfer back to your home school and that you are following your home school’s policies.

Choosing Your Research Project

  • Review Project titles and descriptions below.
  • List 3 (in order of preference) in your Academic Preferences Form, using RSTR as the course code.
  • Program is highly individualized, with limited enrollment.
  • We encourage you to contact Arcadia Abroad Academic Advisors, Leah Cieniawa or Richard Evans II, to discuss your particular research interests further.
Course ID Title Credits Syllabus
RSTR RSLW 392S International Independent Research in STEM Fields 6 PDF

Summer 2026 Research Projects

CAELUS project: supporting the development of a digital twin for supporting a drone network for medical items

Supervisor: Edoardo Patelli

The CAELUS project aims to develop a drone delivery network for delivery medical item and support the emergency service in Scotland (see https://projectcaelus.co.uk). Specifically, this project will contribute to the development of a digital twin, a virtual replica of the physical system, which will be used to simulate and analyse the design and operation of drone network in real-world environments. This project will involve basic coding, analysis different scenarios using the digital twin developed at the University of Strathclyde. The aim of the project is to contribute to the development of the digital twin developed at the university. The main objectives are: 1) getting familiar with Digital Twin technology. 2) perform simulation and scenarios analysis. 3) Contributing to the development of specific modules. Objectives can be refined according the needs and interests of the students.

Relevant Majors: Mechanical Engineering, Computer Engineering, Data Science, Statistics

 

Replacing software user manual with private GTP chat

Supervisor: Edoardo Patelli

Traditional of scientific codes developed by academics are associated with a user manual that is hardly maintained. This often produce a discrepancy between the code and the user manual. In addition, users generally do not use provide manual but they prefer searching the solution on websites or chat-boxes.

This project aims at testing the ability of private GPT to replace user manual on scientific open-source software. This will allow to make scientific software tools more accessible, and free resource required for the maintenance of updated documentation. In turn the code generated by the GPT can also be used for software verification.

The idea is to use the documentation of an open source project names OpenCossan to train a GPT support box. The aim of the project is to identify the feasibility of using generative AI tools (i.e. ChatGTP, Gemini, etc) to replace traditional user manual in software.

Possible objectives are:

  • Identify the most appropriate open-source model to train a local GPT chat
  • Training a local Large Language Model based on documentation of OpenCossan software.
  • Test and verify the capability of the training model

Relevant Majors: Engineering, Computer Science, Data Science

 

Visualization of Fuel Supply Chain Data for Resilience Analysis

Supervisor: Edoardo Patelli

The research focuses on assessing and improving the resilience of fuel and biofuel supply chain infrastructure in Brazil. This project will involve creating visualizations such as maps, graphs, and diagrams to demonstrate how the fuel supply chain is organized and how it has responded to past disruptive events. The main objective is to create visualizations that represent the fuel supply chain and its resilience to disruptions. Among others, these are intermediate objectives.

  • Collect and preprocess data on fuel movements from the Brazilian regulatory authority.
  • Develop maps, graphs, and diagrams that demonstrate the organization of the fuel supply chain and its response to past disruptive events.
  • Use visualization tools to create interactive dashboards displaying key metrics.
  • Prepare a presentation showcasing the visualizations and their insights.

Relevant Majors: Data science, Computer Science, Chemical Engineering, Environmental Engineering

 

Understanding effect of foundation Excavation on Adjacent Subway Structures

Supervisor: Edoardo Patelli

This project focuses on assessing the impacts of foundation excavation on nearby subway structures. As urbanization progresses, construction activities in dense city environments often involve foundation work close to critical infrastructure such as subways. The project will explore how these excavations affect subway tunnels and associated structures, with a specific emphasis on potential deformation, displacement, and long-term stability issues. Through a detailed case study approach, this research will analyse real-world scenarios and apply numerical modelling techniques to evaluate risks and propose mitigation strategies. The primary aim of this project is to investigate the interaction between foundation excavation and adjacent subway structures. The objectives could include:

  • Review the current state of subway infrastructure in relation to ongoing construction activities.
  • Run of numerical modeling (e.g., using software like PLAXIS or FLAC) to simulate different scenarios and understand the effects of excavation on a subway tunnel in different soil conditions.

Relevant Majors: Civil Engineering, Environmental Engineering, Data science

 

Literature Review on Management of Energy Supply Disruptions

Supervisor: Edoardo Patelli

This project will involve conducting a literature review to gather and analyze the experiences of various countries in managing disruptions in the energy supply, particularly for fuels such as diesel, gasoline, biofuels and LPG. The review will explore how governments prepare for and respond to these disruptions, with a particular focus on disruptions caused by climate change-related events This project aims to conduct a comprehensive literature review on the management of energy supply disruptions. The objectives are:

  • Identify and review relevant academic papers, industry reports, and government documents on energy supply disruptions.
  • Collect and summarise the experiences of various countries in dealing with disruptions in the supply of diesel, gasoline, and LPG.
  • Analyse the preparedness measures and response strategies employed by governments to manage these disruptions, including those caused by climate change-related events such as extreme weather, floods, and hurricanes.
  • Prepare a report summarising the key findings and insights from the literature review.

Depending on the quality and extent of the findings, it might be possible to write and submit an academic paper.

Relevant Majors:Engineering, Computer Science, Statistics

 

Stress Testing Brazil's Fuel Supply: Data Analysis of Major Disruptions

Supervisor: Edoardo Patelli

The primary aim is to develop one or more case studies for a supply chain resilience model under development. The student should help to identify and quantify the impact of major past disruptions (e.g., the 2018 truckers' strike, extreme weather events, refinery shutdowns) on Brazil's fuel distribution network. This involves:

  • Researching government reports, news archives and other sources to build a detailed timeline and description of specific disruptive events.
  • Analysing fuel movement data from before, during, and after an event to measure key impacts such as changes in flow volume, delivery delays, and product rerouting.
  • Utilizing the computational models being developed by our research group to assess the model’s usability and to create “what-if” scenarios for crisis preparation and response to increase system resilience.

Relevant Majors: Industrial Systems Engineering, Data Science, Chemical Engineering

 

Software interface for Uncertainty Quantification

Supervisor: Edoardo Patelli

Uncertainty quantification is generally performed by dedication software mostly developed in academia. Over the time we have seen a proliferation of softwares written in different languages (e.g COSSAN software in Matlab, UncertaintyQuantification.jl in Julia, etc.)

The aim of this project is to create a “bridge” between these two projects in order to allows the users familiar with Matlab to take advantage of the new features of the library developed in Julia. This could be obtained by developing wrappers where the syntax of OpenCossan is translated to the systax of UncertaintyQuantification.jl by using some automatic code generation.

The objectives are:

  • Familiarise with software and libraries for uncertainty quantification
  • Develop automatic wrappers from code portability by training and using generative AI tools.
  • Demonstrate the tools by solving provided simple engineering problems.

Relevant Majors: Software Engineering, Computer Science, Data Science

 

AI-driven model for Interval Prediction of Tunnel Displacement

Supervisor: Edoardo Patelli

This project aims to introduce undergraduate students to the application of data-driven methods in geotechnical engineering, focusing on tunnel displacement prediction under urban excavation.

Objectives:

  • Understand how soil and geometrical parameters influence tunnel deformation during excavation.
  • Learn the fundamentals of uncertainty quantification for risk assessment.
  • Develop and test simple predictive models (e.g., neural networks) in Python using provided datasets.
  • Visualize and interpret the results through sensitivity plots and heatmaps.

By the end of the internship, students will produce a short technical report and presentation summarizing the capabilities of the predictive model and its engineering implications

Relevant Majors: Civil Engineering, Data Science, Geotechnical Engineering

 

Dealing with incomplete information with fuzzy variables

Supervisor: Edoardo Patelli

Scientific computations of complex systems are surrounded by various forms of uncertainty, requiring appropriate treatment to maximise the credibility of computations. Current practices face challenges when dealing with empirical information that is often scarce, vague, conflicting and imprecise, requiring expressive uncertainty structures for trustworthy representation, aggregation and propagation.

A framework of uncertain number, a generalised uncertainty representation which unifies probability distributions, intervals, probability boxes, and Dempster Shafer structures has been created for expressive uncertainty representation at different imprecision. This framework, embedded in the Python library pyuncertainnumber, allows for a closed computation ecosystem whereby trustworthy computations can be conducted intrusively or non-intrusively, hence accomplishing faithful management of uncertainty throughout the computational pipeline.

This project will integrate the concept of fuzzy variables/fuzzy distributions into the scope of uncertain numbers and provide a consistent interface in uncertainty management.

Objectives:

  • Engage with the team developing Python library to support its development (e.g. creating an interface between fuzzy distributions and other constructs of uncertain numbers)
  • Testing and validate current implementation
  • Prepare a case study

Relevant Majors: Statistics, Software Engineering, Civil Engineering

 

Physics-Informed Neural Networks for Predicting Physical Related Systems Under Limited Data

Supervisor: Edoardo Patelli
  • Introduce students to physics-informed neural networks (PINNs) and the concept of embedding physical laws into machine learning models.
  • Implement a simple PINN on a benchmark system (e.g., beam deflection, heat conduction, or spring–mass dynamics) using limited data.
  • Evaluate how well the PINN can predict system behavior compared to standard neural networks.
  • Perform basic uncertainty quantification (e.g., Bayesian layers or dropout) to assess model confidence.
  • Summarize findings and provide recommendations on the benefits and limitations of PINNs for small-data problems

Relevant Majors: Data Science, Mechanical Engineering, Computer Science

 

Machine Learning for supporting predictive models

Supervisor: Edoardo Patelli

This project aims to apply simple machine learning (ML) techniques to predict neutron reaction cross sections directly from experimental data. Accurate cross section curves that agree with experiment are essential for reducing uncertainties in reactor simulations, which underpin safe and reliable nuclear power plant design. Instead of relying on complex nuclear models, the student will test fast, data-driven approaches as surrogate models for nuclear data.

The objectives are to:

  • Access and analyse real cross section data (EXFOR).
  • Train basic ML models (KNN, Decision Trees, Neural Networks) using neutron energy as input.
  • Compare predictions with experimental results.
  • Assess how well ML can generalise to unseen data without physics built-in.

By the end, the student will deliver and compare ML models that demonstrate the potential of quick surrogate methods for nuclear data analysis.

Relevant Majors: Physics, Data Science, Statistics

 

Using radars for monitoring structures and rivers

Supervisor: Enrico Tubaldi

This project explores the use of very low-cost Frequency Modulated Continuous Wave (FMCW) radar systems (≈£200 per unit) as a transformative solution for monitoring vibrations of civil structures and surface velocity of rivers. Radars enable remote, contactless measurements of displacements and vibrations with millimetre-level accuracy, providing robust performance in varying environmental conditions and overcoming limitations of optical techniques.

By demonstrating reliable, scalable, and cost-effective radar-based monitoring solutions, the project aims to pave the way for widespread implementation of monitoring solutions, ultimately enhancing the resilience and sustainability of critical infrastructure networks.

Depending on the expertise of the student, the project will focus on different aspects of the research.

Relevant Majors: Computer Science, Software Engineering, Civil Engineering

 

Using cameras for monitoring river flow at bridges

Supervisor: Enrico Tubaldi

Monitoring rivers is essential for understanding hydrological processes, managing flood risk, and ensuring the resilience of surrounding communities. Camera-based systems offer a cost-effective and scalable solution for continuous observation of river flow, water levels, and surface conditions. By applying image-processing and computer vision techniques, valuable hydraulic parameters can be extracted remotely without the need for invasive sensors. This project investigates the use of fixed and mobile camera systems for river monitoring, focusing on methods to enhance data accuracy, automate measurements, and integrate outputs into digital twin and early-warning frameworks.

Depending on the expertise of the student, the project will focus on different aspects of the research.

Relevant Majors: Computer Science, Software Engineering, Civil Engineering

 

From kombucha tea to earthquake resistant soils. Making soils stronger with microbial polymers.

Supervisor: Vernon Phoenix

Overall aim: to stick soil grains together with easy-to-grow polymers grown by microorganisms.

Background and objectives: Microorganisms are able to make a wide range of polymers which can be used in civil engineering applications. One such example is bacterial nano-cellulose. If bacteria can grow these poymers in soils, it could help stick the soil grains together, for example making them more resilient to earthquakes, or suppressing dust during mining operations in dry regions. Fascinatingly, we can make bacterial nanocellulose using the same methods that is used to make Kombucha tea, as the white rind you can get on the top of this tea is nanocellulose. The nanocellulose will be grown using the same microorganisms that are used to make kombucha tea, but these will be grown in reaction vessels containing different soil and aggregate mixes, to explore how the polymers that are produces stick the soil grains together.

Relevant Majors: Civil Engineering, Environmental Engineering, Environmental Science

 

Antimicrobial/antibiotic resistance in the environment

Supervisor: Charles Knapp

Antimicrobial resistance (AMR) is a survival adaptation mechanism. AMR traits can emerge in microorganisms exposed to pollutants employed in interactions with other microbes from different biological kingdoms. In addition, the chemical and ecological community pressure increases the risk of the likelihood that AMR can transfer between microorganisms through mobile genetic elements. The biological mechanisms by which AMR develops under these circumstances are not yet fully understood. In this project, we aim to investigate the linkages between AMR in soil or aquatic systems, with the long-term goal of improving environmental and agricultural well-being. Using aquatic microcosms (surrogate mini-ecosystems), we will examine the role of environmental toxicants (e.g., disinfectants, hydrocarbon pollution or toxic elements--i.e., metals) in the generation and dissemination of AMR.

The nature of the ecosystem and toxicant can be discussed with the project supervisor.

Relevant Majors: Environmental Science, Microbiology, Biology

 

Ecotoxicology: how are environmental pollutants impact microbial ecology

Supervisor: Charles Knapp

You will determine the fate and effects of environmental pollutants in either soil or aquatic microcosms (mini surrogate ecosystems). The toxicant (e.g., disinfectant, metal or hydrocarbon) will be monitored over time, along with the receptor microbial community. Indicators of stress may include disruptions in nutrient cycling, decreases (or shifts) in respiration, population changes, and the emergence of population tolerances.

Multiple students, each with their specialities, will be considered.

Relevant Majors: Environmental Science, Microbiology, Chemistry

 

Hydrological Drivers of Ground deformation in Post-Mining Landscapes

Supervisor: Stella Pytharouli

Aim: Characterise the natural response of the ground above abandoned coal mines

Objectives: using existing available data (1) explore potential relationships between rainfall, soil moisture, hydraulic head and water flow with observed deformations on the ground surface above the mines, (2) characterise the hydrological response of the mine system

Relevant Majors: Data science, Statistics, Civil Engineering

 

Analysis of Flooding Dynamics in Abandoned Coal Mines using Remote Sensing data

Supervisor: Stella Pytharouli

Aim: Explore the flooding process of a coal mine in Scotland (UK) following cease of operations.

Objectives: using available data (1) correlate water levels with observed ground deformation from InSAR data, (2) identify temporal and spatial stages of flooding within the mine

Relevant Majors: Geography, Geology, Civil Engineering

 

Silent Shifts: using AI to detect potential instabilities in abandoned coal mines

Supervisor: Stella Pytharouli

Aim: use of AI algorithms to detect weak seismic signals from shear failures within abandoned mines.

Objectives: using available microseismic data (1) use of AI to automatically detect and locate microseismicity, (2) use locations of microseismic events to identify instability within the mine.

Relevant Majors: Geophysics, Geology, Civil Engineering

 

Enhancing Automated Concrete Defect Detection Using Scene Illumination and AI

Supervisor: Hamish Dow

Identifying defects in concrete structures is vital to ensuring their longevity and resilience. Research to date has focused on spalling and crack detection; other defects like surface corrosion, exposed reinforcement, biofouling and leakage are often overlooked.

Automatic classification of these defects would present a significant cost-, carbon- and risk- reduction for asset managers inspecting tunnels and containments in the transport, defence and energy sectors, as they currently rely on time-/cost- intensive manual inspections.

This internship will investigate how variations in scene illumination (i.e., lighting source and intensity) can be leveraged, in combination with artificial intelligence, to improve the automated detection of concrete corrosion and exposed reinforcement.

Using a bespoke dataset of concrete structures imaged under different wavelengths and lighting conditions, the project will pursue the following objectives:

  • Select an appropriate machine learning or neural network architecture for defect detection.
  • Train a model capable of detecting concrete surface corrosion and exposed reinforcement using the provided dataset.
  • Benchmark the proposed approach against current state-of-the-art methods.
  • Validate the model experimentally on real-world concrete structures.

Relevant Majors: Computer Engineering, Software Engineering, Civil Engineering

 

Synthetic Generation of Concrete Crack Propagation Data Using AI

Supervisor: Hamish Dow

Concrete is the backbone of our built environment, yet it has an unfortunate tendency to crack. Traditionally, detecting these cracks has been the job of a trained inspector, relying solely on the human eye. With recent advances in imaging and artificial intelligence (AI), we can now identify cracks automatically with impressive accuracy.

Where automated crack identification systems still lag behind is in tracking how these cracks change over time, something human inspectors excel at. “Crack propagation” refers to changes in a crack’s length and width. Capturing reliable, high-quality propagation data is both costly and challenging, yet it is exactly what is needed to develop and test the next generation of AI models. At present, our approach has been to strike concrete slabs with a hammer and chisel to gradually worsen the damage. While this is admittedly satisfying, it is far from precise or consistent…

Generative Adversarial Networks (GANs), a class of generative AI models, offer a promising solution by enabling the creation of realistic synthetic crack propagation data. This project will explore how GANs can be trained to produce controlled, high-quality synthetic data to support the development of robust crack detection and monitoring systems.

The objectives of this project are to:

  • Identify and select a suitable GAN architecture for generating synthetic crack propagation data.
  • Develop a workflow for training and validating the GAN using available crack imagery.
  • Evaluate the realism, variability, and usefulness of the generated data for downstream AI training.
  • Compare the synthetic data’s performance against real-world datasets in supporting crack detection models.

Relevant Majors: Computer Engineering, Software Engineering, Civil Engineering

 

AI-Assisted Generation of Infrastructure Inspection Reports

Supervisor: Hamish Dow

Automating the creation of inspection reports has the potential to transform how infrastructure and asset inspections are conducted. While large language models (LLMs) such as GPT can already generate text and interpret images, producing consistent and standards-compliant inspection reports remains a significant challenge.

This project will investigate how multimodal AI systems, capable of processing both visual and textual inputs, can be used to automatically generate inspection reports that align with established engineering inspection codes and formats.

The main project objectives are:

  • Develop a pipeline for automated extraction of inspection insights from images and other input data.
  • Integrate GPT-based models to generate structured inspection narratives and summaries.
  • Ensure compliance of generated reports with relevant inspection standards and reporting codes.
  • Evaluate the consistency, accuracy, and interpretability of AI-generated reports across a range of inspection scenarios.

Relevant Majors: Computer Engineering, Software Engineering, Civil Engineering

 

Illumination-Resistant Image Stitching and Defect Mapping for Infrastructure Inspections

Supervisor: Hamish Dow

Inspection of large concrete structures such as tunnels, bridges, and containment vessels often involves capturing many overlapping images to cover the full surface. Examining these images individually makes it difficult to visualise the overall distribution of surface defects, such as cracks, corrosion, or spalling.

This project will develop an automated image stitching workflow to create high-resolution mosaics of concrete surfaces that are resistant to illumination variability, including shadows, uneven lighting, or external light sources. The mosaics will integrate pre-computed defect maps from an existing dataset, allowing defects to be visualised in their true spatial context.

The main project objectives are:

  • Develop a robust image stitching pipeline capable of aligning and blending overlapping concrete images under varying lighting conditions.
  • Implement feature-based alignment methods, such as SIFT, ORB, or AKAZE, along with blending techniques to minimise visible seams and lighting artefacts.
  • Integrate pre-computed defect maps into the stitched mosaics, ensuring correct spatial placement of defects.
  • Evaluate the quality of stitching, including alignment accuracy, illumination robustness, and correctness of defect positioning.
  • Produce high-resolution mosaics of concrete surfaces with overlaid defect maps for reporting, analysis, or inspection support.

Relevant Majors: Computer Engineering, Software Engineering, Civil Engineering

 

Assessment of aquifer thermal properties for a community Borehole Thermal Energy Storage (BTES) project

Supervisor: Neil Burnside

The Climate Action Through Community Heating (CATCH) community group are working in collaboration with their local Council to deliver a Borehole Thermal Energy Storage (BTES) scheme in support of a low-carbon district heating network (DHN) for a rural town. Initial geological conceptual modelling has revealed the promise of the local shallow sedimentary aquifer for BTES in several locations around the town. Further work is required to better understand the thermal properties of the sandstone and predict its heat storage performance. This project will look to review and capture learnings from other operational BTES systems around the world and assess potential consequences of heat injection in the Kinross area. There are multiple strands of investigation that can either: (i) be included in a single project; or (ii) form different focused studies, allowing multiple students to work on this topic. The overall objectives of the project include:

  • Critically review and assess global examples of BTES schemes including elements such as scale, depth, borehole layout, rock type, rock properties, and heat supply
  • Map and plan ideal location(s) for BTES, taking into account local energy and DHN systems
  • Review knowledge of thermal properties for the Kinesswood Formation sandstone and perform thermal conductivity tests on rock samples using new Strathclyde experimental facility
  • Assess the potential environmental impacts of heat injection and migration on already-stressed local water resources (Loch Leven is facing climate-change related algal bloom challenges)
  • Perform Electrical Resistivity geophysics surveys at identified site(s) to confirm depth to aquifer and water table.

Relevant Majors: Geology, Environmental Engineering, Civil Engineering

 

Mine water geothermal resource assessment

Supervisor: Neil Burnside

Strathclyde has a successful and continually growing research program on mine water geothermal and subsurface heat storage. Active (and future) projects involve various activities such as mapping of mine void spaces, heat resource assessment, geophysical investigations, hydrogeological surveys (including sampling & hydrochemistry), field trials (including pump tests), and modelling different operational and heat end use scenarios. There will be opportunities for student(s) to support this research program in various ways as work at several sites progresses. Specific aims and objectives will depend on the focused topic area of the student and the work available at the time of the summer research effort, e.g. linear and 3D geophysical surveys aim to deliver more accurate imaging of flooded voids in the subsurface, integration of mine abandonment plans aim to deliver robust 3D understanding of mine architecture and optimal locations for drilling, hydrochemical investigations aim to deliver advanced understanding of operational risks and sustainability, energy system modelling will utilise heat end use (e.g., district heating networks, greenhouses) and energy demand / supply scenarios to assess most effective uses for particular locations.

Relevant Majors: Geology, Environmental Engineering, Computer Science

 

Svalbard Hot Springs- what can they tell us about climate change?

Supervisor: Neil Burnside

Major aim is to assess use of arctic hot springs as climate change indicators (and impact assessment tools). We have a recent hydrochemical and stable isotopic dataset from a set of natural thermal springs in Svalbard, and a comparative set of published results from 25 years ago. What can comparison of these data sets, and associated information, tell us about climate change within the Arctic Circle? Can we use records of such prominent natural landmarks to predict climate impacts?

Relevant Majors: Geology, Environmental Engineering, Chemistry