Computer Science and Informatics

Supervisor: Professor Nicholas Caldwell

The Project

Mobile apps can be considered to be in a state of continuous beta release with minor and major releases being issued regularly. As always, changes bring the risk of (re)introducing errors to previously working code, necessitating regression testing to eliminate such unwelcome side effects. Testing user interfaces in particular can frequently be a manual and tedious process, which is at risk of human error due to inattention, missing particular paths, etc.

There are some promising approaches to automating some of this testing, and so this research would seek to investigate artificial intelligence and/or computer vision techniques that can address this problem. We already have one company interested in providing exemplar real-world apps and data sets to support this research. The student will have also access to the University’s new multi-million-pound DigiTech centre, based on the Adastral Park site, which boasts state-of-the-art machines and an AI Compute Server.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or the equivalent). A master’s level qualification would also be advantageous but not essential.

Knowledge of, or a strong interest in mobile app development, software testing and artificial intelligence would be useful.

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Professor Nicholas Caldwell (n.caldwell@uos.ac.uk)

KEYWORDS

Software Testing, Mobile Apps, Artificial Intelligence

Supervisors: Professor Nicholas Caldwell (n.caldwell@uos.ac.uk) and Dr Chris Lewington (c.lewington2@uos.ac.uk )

The Project

The protection of personal data is becoming increasingly important, particularly with respect to the explosion of internet and cloud-based services that make use of personal data as part of a revenue generating business model. The use of (fully) homomorphic encryption techniques has been demonstrated in recent years to be privacy preserving for a growing number of applications, for example gene imputation. The techniques currently require in-depth expert knowledge of cryptography, but recent work (for example at Microsoft) has begun to investigate compiler technologies that would make privacy-preserving encryption accessible to software developers. This research would involve the development of advanced compiler models and software libraries that will then enable service developers to build privacy-preserving tools without needing to know anything about the personal data of an individual. It is expected that one or more cross-discipline case studies will be developed in particular application areas to demonstrate the usability and effectiveness of the research. The student will have access to the University’s new multi-million-pound DigiTech centre, based on the Adastral Park site, which boasts state-of-the-art machines and software dedicated to cyber security.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or the equivalent). A master’s level qualification would also be advantageous but not essential.

Knowledge of cryptography, complier techniques, software development and bioinformatics would be useful.

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Dr Chris Lewington (c.lewington2@uos.ac.uk )

KEYWORDS

Homomorphic encryption, privacy-preserving technology, cloud computing security

Supervisors: Professor Nicholas Caldwell (n.caldwell@uos.ac.uk) and Dr Chris Lewington (c.lewington2@uos.ac.uk )

The Project

The use of Open Source Intelligence (OSINT) to assist in the investigation of alleged crimes is a well-established approach, with organisations such as Bellingcat carrying out ground-breaking work in obtaining and analysing possible evidence from online sources. One of the essential criteria for assessing potential evidence is that of credibility – it is often a non-trivial task, even with dedicated tool support, to sift through large amounts of potentially relevant information to assess the veracity and relevance of one or more datasets. Techniques from artificial intelligence (specifically machine learning) offer promise here, and this research would aim to develop models and corresponding implementations which could help to speed up OSINT analysis and investigation. The student will have access to the University’s new multi-million-pound DigiTech centre, based on the Adastral Park site, which boasts state-of-the-art machines and software dedicated to cyber security.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or the equivalent). A master’s level qualification would also be advantageous but not essential.

Knowledge of open source intelligence techniques, software development and human rights would be useful.

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Dr Chris Lewington (c.lewington2@uos.ac.uk )

KEYWORDS

OSINT, Machine Learning, open source intelligence techniques, passive reconnaissance

Supervisors:
Dr. Kakia Chatsiou, k.chatsiou@uos.ac.uk
Dr. Suha Al-Naimi, s.al-naimi@uos.ac.uk

The Project

Background and Motivation

Messenger RNA–or mRNA– is an essential component of all living organisms. mRNA is a messenger - it interacts with other components in cells, carrying instructions to make a specific protein. mRNA delivers these instructions, or “blueprints” so that cells can put the protein together. Once this process is complete done, an mRNA is broken down by the body.

mRNA teaches human bodies how to make their own medicine – COVID19 mRNA vaccines are only one of the examples where mRNA can be used to give cells directions to make a particular protein. Cardiovascular diseases are another area of application, where studying the proteins missing or produced by such groups of patients can help scientists produce a treatment plan or new medicines.

The structure of mRNA is still not fully mapped and still holds secrets for scientists, and this project explores the idea of using Machine Learning or more specifically, Deep Learning for Natural Language Processing to explore novel ways of extracting patterns and genes and protein information from mRNA sequencing data.

Project Aims

The aim of the project is to work with mRNA sequencing data from cardiovascular disease patients employing methods widely used in detecting patterns in text data from the machine learning and deep learning to study the structure of mRNA and detect its minimal units from existing sequencing data.

The applicant will initially be working on public datasets available in the machine learning and bioinformatics community and inhouse sequencing datasets held by the Suffolk AI (Artificial Intelligence) Research Group members. Machine and Deep learning models will be developed by the student to take multi-modal data as input and generate the high-quality representations as mentioned above. At a later stage, a new dataset would be constructed, and novel algorithms will be developed upon that to answer the challenging questions within this topic and beyond.

Facilities

PhD researchers affiliated with this project will be based in our state-of-the-art DigiTech Centre at Adastral Park, launched in the summer of 2021. A collaboration between the University of Suffolk and BT, with funding from the New Anglia Local Enterprise Partnership (LEP), the centre has been established to provide training in innovative digital skills for people looking to pursue careers in the nationally essential information and communications technology (ICT) sector, as well as fuelling high tech businesses who increasingly require access to a talented technology workforce.

The centre houses four high-end computer laboratories complete with industry-standard software and tools and a dedicated high-end AI server. In addition, PhD students will have access to dedicated comping labs at the Atrium building, in Ipswich, technology focused communal areas, full access to the library resources and other on campus facilities.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours and/or a postgraduate degree with Distinction (or an international equivalent). We also consider applicants from diverse backgrounds that have provided them with equally rich relevant experience and knowledge. Full-time and part-time study modes are available.

Applicants should have the following essential knowledge and skills:

• Strong programming skills (Python, C, Java, etc.)
• Practical experience with deep learning frameworks (e.g., PyTorch and TensorFlow)
• Knowledge of self-supervised learning, CNNs, transformers, geometry, computer graphics rendering models

Applications from applicants with one or more of the following knowledge and skills will be preferred:

• Knowledge of self-supervised learning, CNNs, transformers, geometry, computer graphics rendering models
• Experience with data collection
• Experience with scientific paper writing (e.g., publication or submission)
• Experience with Computer Vision projects
• Experience with Audio or NLP processing
• Strong mathematical knowledge and background

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Dr. Kakia Chatsiou at k.chatsiou@uos.ac.uk

KEYWORDS

Machine Learning, Natural Language Processing, Data Science, Artificial Intelligence, Bioinformatics, cardiovascular diseases, ML for disease detection

Supervisor:  

Dr. Kakia Chatsiou, k.chatsiou@uos.ac.uk

The Project

Background and Motivation

The Industry 4.0 concept aims to the deployment of a large amount of sensor and actuator devices forming an Industrial Internet of Things (IIoT) network. Such IIoT information systems can collect data from all over the shop floor or agricultural unit and can be aggregated to the edge of the network. Edge computing, Edge Intelligence and Federated Machine Learning have been playing an important role in that ecosystem, which heavily rely on the integration of diverse technological solutions to build an intelligent manufacturing ecosystem through the amalgamation of customization, predictive maintenance, digital twin, web-based systems, open connectivity, etc. Edge intelligence, i.e., artificial intelligence (AI) applications with edge computing (edge AI) for training/testing or inference, is an important element for IIoT applications to build a model that can learn from the high amount of aggregated data.

Predictive maintenance is a method of preventing future machinery failure by analysing data during production to detect abnormalities ahead of time, so appropriate measures can be taken before damage can be done to the production line. With faults and failures forecast well before their occurrence, saving money and time for businesses and manufacturing companies, implementing predictive maintenance programs in manufacturing sectors requires condition monitoring of equipment's critical variables and then deploying intelligent techniques to predict the occurrence of failure. The data processing and predictions in predictive maintenance can be ventured as Edge AI techniques, which will be more feasible options for small- and medium-scale industries.

Project Aims

The applicant will work with the supervisor and other academic staff within the Digital Futures and Suffolk AI Research Group to define their specialised area of work depending on their expertise, speaking to one of the areas above. The School and the Digital Futures Institute has excellent links with industry, and it is expected that the project will benefit from direct input and collaborations with local manufacturing partners and cross-pollination of ideas (EDF Sizewell, Openreach BT, Celotex among others).

Facilities

PhD researchers affiliated with this project will be based in our state-of-the-art DigiTech Centre at Adastral Park, launched in the summer of 2021. A collaboration between the University of Suffolk and BT, with funding from the New Anglia Local Enterprise Partnership (LEP), the centre has been established to provide training in cutting-edge digital skills for people looking to pursue careers in the nationally important information and communications technology (ICT) sector, as well as fuelling high tech businesses who increasingly require access to a talented technology workforce.

The centre houses three high-end computer laboratories complete with industry-standard software and tools and a dedicated high-end AI server. In addition, PhD students will have access to dedicated comping labs at the Atrium building, in Ipswich, technology focused communal areas, full access to the library resources and other on campus facilities.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours and/or a postgraduate degree with Distinction (or an international equivalent). We also consider applicants from diverse backgrounds that have provided them with equally rich relevant experience and knowledge. Full-time and part-time study modes are available.

Applicants should have the following essential knowledge and skills:

• Strong programming skills (Python, C, Java, etc.)
• Practical experience with deep learning frameworks (e.g., PyTorch and TensorFlow)
• An understanding of the contemporary issues in Edge computing and AI as applied to the manufacturing or farming industry process (or ability to learn)
• Knowledge of self-supervised learning, CNNs (Convolutional Neural Network), transformers, geometry, computer graphics rendering models

Applications from applicants with one or more of the following knowledge and skills will be preferred:

• Knowledge of self-supervised learning, CNNs, transformers, geometry, computer graphics rendering models
• Experience with working with IoT sensors
• Experience with big data collection & analysis
• Experience with scientific paper writing (e.g. publication or submission to a conference)
• Experience with Computer Vision projects
• Experience with Audio or NLP (Natural Language Processing) processing
• Solid mathematical knowledge and background

Non-UK applicants must meet our English language entry requirements.

Enquiries and Applications

Informal enquiries are encouraged and should be directed to Dr. Kakia Chatsiou at k.chatsiou@uos.ac.uk

KEYWORDS

Machine Learning, Data Science, Artificial Intelligence, Manufacturing AI, Agriculture AI, ML (Machine Learning) for Embedded Systems and IoT (Internet of Things), Edge Artificial Intelligence, Industrial Internet of Things, Federated Active Transfer Learning

Supervisor: Adam Clayden a.clayden@uos.ac.uk

The Project

Previous research into shared participation in recreational experiences has shown to positively contribute to factors such as wellbeing and experience enjoyment (Boothby, et al., 2014; Boothby, et al., 2016). Additionally, there have been recent exploratory studies aimed at better understanding player perspectives with regards to shared recreation (Haqq & McCrickard, 2020). What is not currently clear, are the underlying mechanisms that contribute to a gameplay experience that is not only enjoyable, but also promotes socially meaningful interactions and other factors such as connectedness with nature.

The goal of this project, is to make contributions to the exergame and interdependent play space for the purpose of improving a person’s physical and mental wellbeing. It is expected that the PhD student will engage in a body of research aimed at understanding and establishing mechanisms that contribute to this area of game design and also design and build such games that can be played by the general public. Typical outcomes will involve submitting theoretical and practical work to annual conferences such as Foundations of Digital Games and CHI PLAY. At the University of Suffolk, you will have access to our state-of-the-art games lab, the specialist laboratories in our DigiTech Centre, as well as other relevant facilities such as our Sports Science lab.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or the equivalent). A master’s level qualification would also be advantageous but not essential.

Knowledge of games development and/or programming is essential. It is advantageous to have a working knowledge of univariate statistics but this is not essential and can be learned throughout the PhD.

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Adam Clayden (a.clayden@uos.ac.uk)

References

Boothby, E. J., Clark, M. S., Bargh, J. A., (2014) ‘Shared experiences are amplified.’, Psychological Science, 25(12), 2209-2216.

Boothby, E. J., Smith, L. K., Clark, M. S., Bargh, J. A., (2016) ‘Psychological distance moderates the amplification of shared experience. Personality and Social Psychology Bulletin, 42(10), 1431-1444

Haqq, D., McCrickard, D. S., (2020) ‘Playing Together while Apart: Exploring Asymmetric and Interdependent Games for Remote Play’ Extended Abstracts of the 2020 Annual Symposium on Computer-Human Interaction in Play (CHI PLAY ’20 ESA). Virtual Event, Canada 2-4 November 2020. ACM, New York, NY USA, pp. 253-256

KEYWORDS

Exergaming, HCI, interdependent, share recreation, games design

Supervisors:

Professor Nicholas Caldwell (n.caldwell@uos.ac.uk) and Dr Chris Lewington (c.lewington2@uos.ac.uk )

About the project

The School of Technology, Business and Arts at the University of Suffolk is inviting applications for the following project.

The Project

AI systems are frequently portrayed as being complex, unpredictable and often prone to ‘hallucination’. Treating the development of an AI system as an inherent engineering problem, the current products, services and literature point to a significant absence of software engineering best practice. The title of the project is deliberately derived from Dijkstra’s seminal book ‘A Discipline of Programming’, which advocates for a structured and rigorous approach to software development. Arguably such discipline is missing from the current ways in which Machine Learning (ML) systems are currently designed and built.

This research would focus on the development of sound engineering principles and practices to support the development of robust, testable and maintainable AI systems, using as a foundation the core principles and experience generated from the many decades of software engineering best practice. For example, models, techniques and tools developed for best practice in software architecture, design techniques using Data Flow Diagrams, Threat modelling techniques, best practice programming techniques and testing frameworks at all levels will be ‘repurposed’ to create bespoke equivalents that put AI systems development on a much stronger engineering footing, bringing benefits of traceability, maintainability and understanding that is currently missing. Tools to support Intelligent Machine Engineering will also be produced, for example design tools and IDE plugins.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or the equivalent). A master’s level qualification would also be advantageous but not essential.

Knowledge of Machine Learning techniques and libraries, software engineering principles and IT systems development best practice would be useful.

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Dr Chris Lewington (c.lewington2@uos.ac.uk )

KEYWORDS

Machine learning, software engineering, software lifecycle, software architecture and design

Supervisors:

Professor Nicholas Caldwell (n.caldwell@uos.ac.uk) and Dr Chris Lewington (c.lewington2@uos.ac.uk )

The Project

The classical perimeter-based approach to cyber security, which focuses most attention on securing a network boundary, is becoming less appropriate particularly for cloud-based architectures, where users, applications, servers, and access devices can be globally dispersed and varied. Moreover, a perimeter-based security solution will not guard against insider threats, as users or devices located within the perimeter are assumed to be ‘safe’. The approach of Zero Trust Architectures (ZTA) instead places emphasis on users, attributes, and resources, assuming instead that every type of access is untrusted and the mantra of ‘never trust, always verify’ is applied throughout a system. Frameworks exist for ZTA, however actual implementations are few in number for various reasons.

Related to the lack of concrete ZTA examples is the fact that even those that exist rely upon standard cryptographic techniques for authentication and access control. There is much interest currently in quantum technologies, and it has already been shown that (with a sufficiently powerful quantum computer) existing public-key cryptographic techniques, which secure the whole of the internet, could be broken easily. ZTA solutions, therefore, will need to address this issue to ensure that future implementations will be quantum-robust, and currently there are several Post-Quantum Cryptography (PQC) solutions that NIST are in the process of finalising as standards for digital signatures and key encapsulation/exchange.

This PhD work will build on research that is currently under way at the University of Suffolk to create a robust ZTA example implementation that can also be suitably modified to ensure post-quantum safety.

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or the equivalent). A master’s level qualification would also be advantageous but not essential.

Knowledge of Zero Trust Architectures (ZTA), attribute-based/functional encryption, software engineering and system design, cryptography (especially post-quantum techniques) would be useful.

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Dr Chris Lewington (c.lewington2@uos.ac.uk )

KEYWORDS

Zero-trust, quantum cryptography, attribute-based encryption, cloud computing security

Supervisors:

Professor Nicholas Caldwell (n.caldwell@uos.ac.uk)

The Project

Computing systems and software can and does aid in addressing the United Nations Sustainable Development Goals through reducing travel (communicate via video conferencing rather than commuting), calculating fuel-efficient logistics, smart agriculture, and so on. All of these are the use of computing for sustainability. Unfortunately the manufacture, operation and disposal of computing resources has a large environmental cost and as data centres and AI applications have grown in popularity, so has the contribution of the information and computing technology sector to total greenhouse gas emissions. By 2018, it was estimated to have reached 4 to 5% (twice the emissions from air travel). Bitcoin/cryptocurrencies and modern data centers have been reported as placing significant strain on national energy infrastructures. This research is therefore focused on the sustainability of computing and what techniques can be employed to reduce energy consumption and hence the carbon/greenhouse gas footprint of computing use, specifically software.

In previous research, we have found intriguing and unexpected insights in the relationships of energy usage versus performance, task complexity, and execution speed for real software components. We would like to expand our understanding of the factors underpinning energy consumption of different type of software components and especially where software components have unexpectedly high performance and energy costs. The end goals of this research will be how to determine whether a particular software component is “sustainable” and how we can develop software that is more energy efficient in terms of programming and software engineering

Candidate Requirements

Applicants should hold, or expect to receive, a First Class or good Upper Second Class Honours degree (or the equivalent). A master’s level qualification would also be advantageous but not essential.

Knowledge of computer programming, software engineering principles and IT systems development best practice would be useful.

Non-UK applicants must meet our English language entry requirement.

Enquiries and Applications

Informal enquiries are welcomed and should be directed to Professor Nicholas Caldwell (n.caldwell@uos.ac.uk)

KEYWORDS

Sustainable computing, software engineering, software lifecycle, software architecture and design