AI-Enabled Research
AI will play a crucial role in enabling and accelerating research across all disciplines. We will transform research by providing new methods for understanding a wide range of data, detecting key or rare events in massive data sets, and extracting new insights in scientific and medical data.
Powering-up Research
We believe that AI has the potential to massively accelerate our progress towards solutions for the big global challenges, such as those tackled by our flagship institutes and centres, in health and social care, sustainable food, manufacturing and climate science.
More broadly we believe that AI can transform research across the disciplines: in physics, chemistry, biosciences, medicine, the arts, social science and engineering, leading to greater insights and discovery and ensuring Sheffield maintains its position as one of the UK's top research Universities.
However, moving to this new age of AI-enabled research is not straightforward. AI's interdisciplinary nature and the inherent complexity and variation in typical research data present real challenges that need to be overcome.
To help address these challenges, the CMI and the CMI's AI Research Engineering (AIRE) team will focus on the following questions:
-
How do we bring AI knowledge, tools and expertise to different application areas, especially those which are currently lacking in AI expertise and skills, so that AI can be more widely embedded in research across all disciplines?
-
How can we better exploit and integrate the wide range of research data to achieve novel research insights - (data is wide ranging both in terms of modality, e.g. images, sounds, speech, numeric, text, and in terms of the application area, e.g. health and medicine, physics, chemistry, biosciences, engineering, the arts and social sciences)?
AI-enabled research at the University of Sheffield
There are various initiatives currently working to progress AI-enabled research at Sheffield.
AI-enabled research in our institutes, centres and research groups:
The UKRI Centre for Doctoral Training (CDT) in Speech and Language Technologies (SLTS) is a world-leading hub for training scientists and engineers in speech and natural language processing (NLP / NLProc) – two core areas within artificial intelligence (AI) which are experiencing unprecedented growth and will continue to do so over the next decade.
The Insigneo Institute - use healthcare data and Artificial Intelligence (AI) to speed up accurate diagnosis of disease, help work out what treatments work for what diseases, and understand how to predict who will get disease.
Healthy Lifespan Institute (HELSI) - are identifying the determinants and predictors of multimorbidity and frailty (including socio-economic and environmental factors and biological processes) using Artificial Intelligence approaches.
The Artificial Intelligence and Computational Neuroscience Group - The AICN Group is a multi-disciplinary research group which aims to combine innovative and original computational methods with state-of-the-art brain imaging techniques in understanding, detecting and developing treatments for neurological and psychiatric conditions.
Neuroscience - researchers from the department of Neuroscience have developed a digital tool (CognoSpeak) that uses artificial intelligence and speech technology to automatically analyse language and speech patterns that could warrant further specialist investigation and be early signs of dementia or Alzheimer’s disease.
Chemoinformatics - AI/Machine learning for drug discovery is a key research theme in the Chemoinformatics research group. They have applied AI techniques to categorise and prioritise chemical reactions to aid in chemical synthesis and have analysed the uncertainty of the predictions of machine learning models using error models and conformal prediction.
The Advanced Manufacturing Research Centre (AMRC) - AI and other data science techniques play a pivotal role in subtractive machining research and are critical parts of their digital pathway. Coupled with digital twins, these are digital building blocks towards autonomous machining, helping deliver on two fundamental goals: efficiency and productivity. Data Science and AI capabilities encompass the whole data pipeline and are exploited through end applications such as augmented intelligence, machine learning operations practices, integrated computer vision and other areas including natural language processing, through-life analysis and comprehensive data-driven insights.
The Digital Humanities Institute - the UK's leading centre for digital humanities, supports the innovative use of technology and computation in arts and humanities research as both a method of inquiry and a means of dissemination.
Department of Civil and Structural Engineering - researchers are applying digital design techniques to: reduce environmental impacts of designs at the outset; manage buildings and infrastructure throughout their lifespan for effective building maintenance and material reuse; optimising structures to improve resilience, durability, and efficiency and reduce resource consumption. Researchers are also exploiting data-driven technologies including urban sensing to understand everything from air pollution to what our cities are made of. Using techniques such as photogrammetry and multi-spectral imaging, they are building a digital twin of Sheffield including live sensing to develop a dynamic understanding of city systems.
Various research groups at Sheffield are developing machine learning and foundational AI technologies:
Machine Learning group - explores and develop the capacity for algorithms to learn and make decisions and predictions from their environment. We follow a series of complementary approaches within the group, from biologically inspired computational models to probabilistic modelling and dimensionality reduction.
Computer Vision Research Group - dedicated to pushing the boundaries of computer vision, computer graphics, animation, and complex systems simulation through ground-breaking research.
Speech and hearing group - concerned with computational modelling of auditory and speech perception in humans and machines; robustness in speech recognition and large vocabulary speech recognition systems and their applications.
Natural Language Processing group - one of the largest and most successful language processing groups in the UK with a strong global reputation. Research Themes include: NLP for social media, Information Access, Language Resources and Architectures for NLP, Machine Translation, Human-Computer Dialogue Systems, Detection of Reuse and Anomaly, Foundational Topics, Biomedical Text Processing.
Security of Advanced Systems group - carries out fundamental and applied research in cybersecurity, drawing on expertise in correctness by design and in machine learning and are building wider collaborations with high technology domains.
Pervasive Computing - focused on co-designing, developing, applying and evaluating pervasive technologies in applications that integrate with people’s lives, work, and leisure activities. The group’s research covers, but is not limited to, four main themes: Mobility Analysis, Pervasive Healthcare, Digital Agriculture, Co-Design and Inclusivity.
Testing group - one of the largest groups of its kind in the UK developing innovative approaches to software testing and quality assurance. Themes the group investigates include: testing for AI-enabled cyber physical systems, such as automated driving systems; testing for robots and autonomous systems, model-based testing, search-based testing, security testing, reverse engineering, model-driven engineering, multi-agent modelling and XML data processing.
Complex Systems and Signal Processing Research Group - internationally leading in the development of techniques and algorithms for complex systems analysis, control and signal processing and the application of these in emerging areas of science, engineering and medicine. The group is renowned for its work on the identification and analysis of complex spatio-temporal systems, nonlinear signal processing, and the analysis and design of nonlinear systems in the frequency domain.
Intelligent Systems, Decision and Control research group - internationally renowned for its work on multi-objective evolutionary optimization algorithms, intelligent health monitoring and fault diagnosis, decision support systems for biomedicine, information processing and computational data modelling.
Dynamics research group - conducts research into conducts research into vibrations, dynamics of structures and acoustics. Their research topics include nonlinear structural dynamics, digital twins, structural health monitoring, active and passive vibration control and acoustics.