Now, Manchester researchers are turning to physics-informed AI, to model how the global carbon cycle behaves and to test the potential of different carbon removal strategies virtually. Their approach offers more flexible predictions than traditional numerical models and can estimate uncertainties where data is missing.

Crucially, these AI models also deliver results far faster and with a lower computational burden. As project lead Dr Peyman Babakhani explains: “Field experiments, especially for ideas such as ocean fertilisation, are costly and slow. With the advent of physics-informed AI, we can replace or facilitate such experimental campaigns with predictive models that can incorporate a more accurate representation of physical processes than common numerical models and are also faster. This enables us to study proposed CO₂ removal methods at scale.”

One theory the team are exploring is the use of engineered nanoparticles to make ocean fertilisation more effective. Studies suggest that nanoparticles of iron, silica or aluminium could boost plankton growth, extend bloom lifetimes and increase the amount of carbon that sinks. However, such methods carry their own costs and risks that need to be evaluated in the laboratory and in silico before field trials.

The AI-powered models might soon be the key to testing ambitious climate solutions before we take them into the real world, helping us to more efficiently combat climate change.

Meet the researcher

Dr Peyman Babakhani is a lecturer of Geo-environmental Engineering in Manchester’s Department of Civil Engineering and Management. His research uses nanotechnology to address environmental issues, such as climate change and water pollution. He uses various techniques, such as physics-informed artificial intelligence and numerical mass- and population-balance models, to model different environmental scales ranging from nano to global. He is a member of Forum which focuses on ocean fertilisation as a CO2 removal approach.

• Email Dr Babakhani >>
• Email the Business Engagement Team >>

Read his papers

A team consisting of Nathan Shankar, Professor Hujun Yin and Dr Pawel Ladosz from The University of Manchester is tackling this challenge by teaching robots to ‘see’ in the dark. Their approach uses machine learning to reconstruct clear images from infrared cameras – sensors that can ‘see’ even when no visible light is present.

The breakthrough means that robots can continue using their existing vision algorithms without making changes, reducing both computational costs and the time it takes to deploy them in the field.

As project lead Dr Pawel Ladosz explains: “Our work enables robots to function in darkness with minimal changes to their platforms. This lowers development costs, speeds up deployment and opens the door to operations in some of the most challenging environments imaginable.”

Looking ahead, the team sees potential beyond low light settings. By adapting their system to sensors such as sonar or thermal cameras, they could potentially expand robotic vision into an even wider range of extreme conditions.

Meet the researcher

Dr Pawel Ladosz is a Lecturer in Engineering Systems for Robotics in the Department of Mechanical and Aerospace Engineering. His research interests lie in making robots more autonomous using vision-based sensors, and he has extensive experience with aerial, ground-based and underwater mobile robots. Dr Ladosz’s most recent research includes reinforcement learning, visual SLAM, heterogeneous robotic teams and supervised machine learning.

• Email Dr Pawel >>
• Email the Business Engagement Team >>

Read his papers

That’s the challenge at the heart of work being undertaken by Dr Mehdi Hellou as part of PRIMI, an EU-funded project exploring how robots can develop a ‘theory of mind’ – the ability to infer what people believe, prefer or intend. The aim is to develop autonomous technologies that might anticipate when someone needs help, adapt their behaviours over time, or respond to situations in a more socially intelligent way.

To achieve this, researchers are drawing on insights from psychology, neuroscience and artificial intelligence to create robots that combine motor intelligence (how they move), with cognitive intelligence (how they reason).

As project lead Dr Hellou explains: “It’s important to develop autonomous systems that can assist humans in their daily life, but also in critical scenarios such as healthcare or nuclear waste decommissioning. This requires machines capable of adapting their behaviours to different users and environments.”

The project’s vision will be tested in clinical pilot studies on stroke rehabilitation, where humanoid robots could support patients’ recovery.

If successful, PRIMI could help to usher in a new generation of socially aware robots that are not only more capable of learning in real time, but also more relatable and trustworthy.

Meet the researcher

Dr Mehdi Hellou is a Research Associate in Artificial Intelligence and Robotics at Manchester’s Centre for Robotics and AI. He previously completed his PhD in Robotics and AI under an EU-funded project called PERSEO, which looked at enhancing the cognitive abilities of social robots by using ‘Theory of Mind’.

• Email Dr Mehdi>>
• Email the Business Engagement Team >>

Read his papers

Researchers at The University of Manchester are developing AI-based methods to map how firms adapt to supply chain risks. By analysing data from international firms, including site visit transcripts, the team uses large language models to detect where and why networks are changing – from concentration around specific suppliers to diversification across regions.

This research offers a new lens on strategic management, showing how companies respond to uncertainty and external shocks. Insights from the project could inform policy and industry efforts to build more transparent, secure, and sustainable supply chains.

Linyi Guo, the PhD researcher leading this project explains: “I believe innovation should be inclusive and driven by real-world needs, especially in supply chain transparency and corporate strategy. By combining AI with strategic analysis, we can uncover how global networks evolve – helping businesses and policymakers make better, fairer decisions in complex systems.”

Meet the researcher

Linyi Guo is a PhD researcher in Science, Technology, and Innovation Policy, based at the Alliance Manchester Business School. Under the supervision of Professor Andrew James and Professor Kieron Flanagan, her primary research interests are innovation management and innovation policy, with a focus on high-tech industries. Her expertise includes Python, SPSS and MySql.

• Email Dr Linyi >>
• Email the Business Engagement Team >>

Read her papers

The research work led by Dr Arijit De at the Alliance Manchester Business School, are using advanced artificial intelligence to bring clarity to this turbulent market. Their MarineFuelAI system combines historical fuel data, economic indicators and port-specific variables to forecast bunker fuel prices for different fuel grades at several global ports, for up to 60 days in advance.

The technology doesn’t just crunch numbers. Enhanced with explainable AI techniques, it can reveal the hidden drivers behind fuel price movements, from regional demand shifts to geopolitical events like the Russia–Ukraine conflict.

These tailored, route-based forecasts can give shipping companies much more confidence in their refuelling decisions. As Dr Arijit De explains: “This approach brings clarity around future fuel prices, cuts bunkering costs and helps global shipping sail confidently through uncertainty, toward a more efficient, resilient and future-ready industry."

By reducing both risk and expense, improving operational efficiency, MarineFuelAI could help the maritime sector navigate fuel volatility while supporting a more sustainable global shipping industry into the future.

Meet the researcher

Dr Arijit De is an Associate Professor at the Alliance Manchester Business School, a Chartered Fellow of CILT and an Industrial Engineer (MTech, PhD). He applies AI, machine learning, intelligent algorithms and optimisation to real-world challenges on freight and maritime logistics, supply chain management and sustainable operations. His research is funded by Horizon Europe, ESRC, Department for Transport, EPSRC, Innovate UK and published in leading operations and transportation journals.

• Email Dr Arijit >>
• Email the Business Engagement Team >>

Read his papers

Through diary studies and interviews with fashion professionals already using AI in design and product development, the project examines how emerging tools such as digital prototyping and generative design can reduce physical sampling, improve material selection, and enable more circular production cycles.

This human-centred approach reveals both the potential and the practical barriers to adopting AI in creative workflows, offering insight into how technology can support a just transition to sustainable, data-driven fashion.

Dr. Courtney Chrimes, Lecturer in Digital Fashion Marketing explains: “By rethinking design through AI and circularity, we can transform fashion from one of the world’s most wasteful industries into a force for regenerative change.”

By bridging creativity and computation, this research positions Manchester at the forefront of sustainable innovation – helping an industry long associated with excess move toward circular, intelligent design.

Meet the researcher

Dr. Courtney Chrimes is a Lecturer in Digital Fashion Marketing at The University of Manchester. Her research explores how industry 5.0 technologies, specifically AI, can support sustainable fashion, aligning with UN SDGs 9 & 12. She co-founded the AI in Fashion Consortium and leads projects on AI-driven decision-making and material selection, with work published in top peer-reviewed journals.

• Email Dr Courtney >>
• Email the Business Engagement Team >>

Read her papers

A team of Manchester researchers have been exploring this issue through in-depth research with accountancy firm Beever Struthers, looking at the use of generative AI through in-person observation, chat logs and interviews. Their early findings reveal that whilst AI is highly effective at streamlining repetitive tasks, if firms aren’t careful it could also start to encroach on areas where human expertise is essential. For example, AI-generated summaries may speed things up but risk losing crucial context, whilst the technology’s ability to make assumptions could blur lines of professional responsibility.

The team’s study highlights that accounting relies on more than technical analysis; client interactions, on-site fieldwork and mentoring are vital to developing the professional judgement that underpins trust in the field. These are skills AI cannot currently replicate.

Led by Dr Sung Hwan Chai, Professor Brian Nicholson and Dr Leonid Sokolovskyy the project aims to redefine what professional judgement means in an AI-enabled world, offering a framework that could help accountants to use generative AI responsibly, and harnessing its efficiencies while protecting the human insight that makes their work reliable.

Dr Chai explains: “Our project has both academic and practical impact. First, we’re redefining what “professional judgement” means in accounting – in a way that applies to all areas of the profession, not just auditing, and takes new technologies like AI into account. Second, we’re working with the Institute of Chartered Accountants of Scotland (ICAS) to create a report that will help accountants use generative AI responsibly and ethically in their work.”

Meet the researcher

Dr Sung Hwan Chai is a Lecturer in Accounting in the Accounting and Finance division of the Alliance Manchester Business School. He specialises in interdisciplinary research between management accounting and information systems, using a qualitative case study approach. His research interests are in the impact of current and future technologies on management accounting practices, such as performance measurement and management, surveillance and information communication practices.

• Email Dr Sung Hwan >>
• Email the Business Engagement Team >>

Read his papers

Researchers at The University of Manchester have created SylloBio-NLI, a first-of-its-kind framework that systematically tests the logical reasoning ability of AI models.

Using examples similar to classic syllogisms – “All men are mortal. Socrates is a man. Therefore, Socrates is mortal.” – the team adapted this structure to biomedical data to reveal where models succeed and where they fail.

Their findings show that while AI can make intuitive connections, even advanced open-source models struggle with consistent logical reasoning when applied to biomedical problems. By quantifying these limitations, the research provides critical evidence for the safe use of AI in scientific discovery and clinical decision-making.

Danilo Carvalho, Principal Clinical Informatician for the Digital Cancer Research team at the National Biomarker Centre, within Cancer Research UK Manchester Institute explains: “By exposing where AI reasoning breaks down, we can build systems that support biomedical research with certain scientific evidence guarantees.”

The team’s open-access methodology offers a vital tool for improving the transparency, reliability, and future design of AI technologies used in medicine, supporting Manchester’s commitment to ensuring responsible AI and digital health innovation.

Meet the researcher

Dr Danilo Carvalho is a Principal Clinical Informatician for the Digital Cancer Research team at the National Biomarker Centre – . He is qualified as a Computer and Information Scientist (MSc, PhD) and is an expert in explainable and controllable mechanisms for representation learning, which is the building of computer-based numerical models of physical or abstract reality, from the meaning of words to gene interactions.

• Email Dr Danilo >>
• Email the Business Engagement Team >>

Read his papers

To tackle this challenge, a Manchester research team led by Dr Danilo S. Carvalho and Dr André Freitas have developed new software frameworks – LangVAE and LangSpace – that significantly reduces both hardware and energy resource needs for controlling and testing LLMs to build explainable AI.

Their technique builds compressed language representations from LLMs, making it possible to interpret and control these models using geometric methods (essentially treating the model’s internal language patterns as points and shapes in space that can be measured, compared and adjusted), without altering the models themselves. Crucially, their approach reduces computer resource usage by over 90% compared with previous techniques.

This leap in efficiency lowers the barriers to entry for developing explainable and controllable AI, opening the door for more researchers, startups and industry teams to explore how these powerful systems work.

Dr Carvalho explains, “We have significantly lowered entry barriers for development and experimentation of explainable and controllable AI models and also hope to reduce the environmental impact of these research efforts.

“Our vision is to accelerate the development of trustable and reliable AI for mission-critical applications, such as healthcare.”

Meet the researcher

Dr Danilo Carvalho is a Principal Clinical Informatician for the Digital Cancer Research team at the National Biomarker Centre – . He is qualified as a Computer and Information Scientist (MSc, PhD), and an expert in explainable and controllable mechanisms for Representation Learning: the building of computer-based numerical models of physical or abstract reality, from the meaning of words to gene interactions.

• Email Dr Danilo >>
• Email the Business Engagement Team >>

Read his papers

Now, researchers at The University of Manchester working with Unilever, have created an AI-powered ‘self-driving’ laboratory that promises to change the way chemical innovation happens. 

Their new system uses physics-guided AI and is designed to learn efficiently by choosing only the most valuable experiments, cutting down the number of tests needed to reach reliable results. Instead of endlessly tweaking variables in the lab, it learns from every outcome, refining its models to predict what will work best next.  

This not only saves time and resources but delivers deeper insights into the underlying science. The resulting chemical processes can be developed faster, scaled more efficiently and designed with sustainability in mind – from cleaner consumer goods to greener manufacturing systems. 

As project lead Dr Dongda Zhang explains: “By advancing AI-powered self-driving labs, we help our industrial partners enhance digital maturity, embrace culture change and accelerate sustainable innovation – driving smarter, faster and cleaner manufacturing that benefits both industry and society.” 

Traditionally, discovering these materials has been a slow process of trial and error in the lab. Now, a team at The University of Manchester, led by Dr Qian Yang, are using artificial intelligence to speed things up. 

Their system doesn’t just crunch data – it learns the way physicists think about materials, helping to predict which ones are worth making and testing and even guiding their design and manufacturing. This way, AI acts less like a passive tool and more like an active ‘lab mate’, working alongside the researchers to unlock smarter and faster innovation. 

As Dr Yang explains: “Our work is building the materials foundation for the future. From lossless superconductors to clean energy catalysts, new quantum materials will underpin the next wave of scientific and technological progress.”

Now, researchers at The University of Manchester are taking this a step further. Led by Dr Jie Wang, the team have developed an AI-powered ‘digital twin’ of the human vascular system that makes patient-specific blood flow simulations faster, more accurate and more accessible for clinical use. 

By combining advanced computer models with machine learning, their system can quickly predict key indicators such as pressure and wall shear stress – measures that help doctors understand how blood moves and where risks may lie. Unlike traditional methods that require time-intensive, high-powered computing, this approach runs efficiently on the open-source AortaCFD app, paving the way for real-time, personalised insights. 

As Dr Wang explains, their long-term goal is to give clinicians a powerful tool for precision care: “Our work uses advanced computer modelling to create realistic “digital twins” of patients’ blood flow, allowing doctors to simulate and predict how blood moves through the body quickly and accurately.  

“By combining different modelling techniques, the research turns complex simulations into practical tools that help clinicians plan personalised treatments.  

“Future work will use AI methods, such as physics-informed neural networks (PINNs), to make these predictions even faster and more precise.” 

Now, researchers led by Dr Tim Tang at The University of Manchester are using artificial intelligence to unlock fresh insights. By training AI on computer simulations that mimic the ocean in fine detail, the team have uncovered a new mathematical equation that describes when and how waves break. 

Unlike traditional ‘black box’ AI, which makes predictions without showing its reasoning, this method provides interpretable results by human beings. It has revealed that, in deep water, the wave breaking can start from a speed difference between the surface and bulk of water underneath, which creates a small “waterslide”. Similar to landslide, the water at surface running down is likely to trigger a full energic breaking. 

The new equation could help scientists model the ocean more accurately and simulate breaking waves more efficiently, with applications ranging from advancing offshore renewable engineering to climate forecasting. As Dr Tang explains: “Wave breaking is one of the last unsolved puzzles in ocean science. By combining AI with physics, we’re not just improving our models – we’re uncovering new insights about the ocean wave itself.”

To address this challenge team, led by Dr Mahdi Saleh, have designed a deep learning model that ‘cleans’ brain signals (EEG) in real time, running on small, low-power wearable devices. By filtering out unwanted noise and interference, the technology delivers clearer, more reliable readings without the need to send sensitive brain data to remote servers. 

This breakthrough is the first to show that advanced AI for EEG monitoring can work on compact, wearable devices. Whilst each device offers trade-offs between speed and power consumption, the research proves that real-time, portable and secure brain monitoring is possible. 

As Dr Saleh explains: “This turns hospital-level monitoring into an everyday reality, helping people with epilepsy now and unlocking a future of smarter healthcare.” 

Beyond epilepsy care, the innovation could pave the way for next-generation health devices and brain-computer interfaces, where privacy, portability and real-time performance are essential. 

This work was supported by the EIC Pathfinder RELIEVE Project. 

Her research focuses on designing realistic digital materials – like cloth, hair, and fur – that behave naturally under changing light and movement. Traditionally, achieving this realism meant simulating every bounce of light, a process that demands huge amounts of time and computing power. Instead, Dr Montazeri’s approach uses AI to learn how light interacts with different materials, creating lifelike textures far more efficiently. The result is digital content that looks and feels real but can be generated in a fraction of the time. 

“As digital experiences become more immersive, how real things look matters more than ever,” she says. “By combining AI with advanced rendering techniques, we can make virtual worlds more convincing and engaging for everyone.”

At The University of Manchester, researchers are using artificial intelligence to bring welding into the digital age. Their project combines advanced computer simulations with machine learning models that can optimise welding outcomes and spot stresses that could occur during the process. 

Instead of relying on trial-and-error, engineers and welders can now explore designs virtually, identifying potential flaws before they happen and ensuring stronger, safer structures over time. 

Through the team’s approach, what once required hours of detailed simulation or physical testing can now be assessed in moments. 

As researcher Zeyuan Miao explains: “By combining advanced simulations with surrogate machine learning models, we automate and accelerate welding process design. Our approach reduces trial-and-error, minimises defects, and shortens development time – bringing intelligent decision-making directly to the welder. 

“Welding is everywhere – now we’re turning it into smart science. With AI-driven insight, we can help welders worldwide build safer structures, faster and with confidence.” 

Their project combines large-scale data analysis with on-the-ground case studies from labs across the UK and internationally. By analysing millions of publications in databases such as OpenAlex, the team can track how scientists apply AI, including how generative AI tools, such as ChatGPT, are spreading through scientific fields.  

Their early findings on generative AI show that whilst the US and China lead in overall volume of scientific papers, smaller research economies are also embracing the technology, often with significant results. Interestingly, research teams working on generative AI tend to be slightly smaller than those in other AI fields, suggesting a different style of collaboration is emerging.  

But rapid adoption also brings challenges. Summarising documents or generating code with AI can accelerate research, yet it raises questions about responsibility, governance and the line between human and machine judgement. 

Professor Cornelia Lawson, Professor of Economics of Science and Innovation, explains: “This project probes how AI shapes scientific discovery and how it can be used responsibly, creatively and equitably to benefit researchers and society alike.” 

Her colleague, Professor Philip Shapira, Turing Fellow and Professor of Innovation Management and Policy, Manchester Institute of Innovation Research, adds: “AI is reframing science, changing skill demands, influencing collaboration and transforming opportunities. Yet, AI’s impacts on scientific novelty and creativity are uncertain – a knowledge gap that our project is now focusing on.”  

By understanding AI's impacts in science, this research will help shape future research and innovation strategies, competitiveness, knowledge advancement, responsibility, and societal implications. 

Interested in presenting your work at The Digital Environment Conference 2026, hosted at SISTER on 1st April 2026? We are looking for individuals to present their research in 15 minute speakers slots, or present their work on a poster board at the event.

Please email Jade at digitalfutures@manchester.ac.uk with your presentation and/ or poster title, and topic or area of research. 

Please note that the open call for presentation or poster submissions deadline is Friday 27th February 2026.

In 1950, Alan Turing published “Computing Machinery and Intelligence”, one of the first papers on artificial intelligence. But theory alone wasn’t enough. AI needed powerful computing. And at Manchester, that power was being built. 

In 1948, Frederic C Williams, Tom Kilburn and Geoff Tootill created the Manchester Baby – the first stored-program computer.   

Three years later, the Ferranti Mark I was unveiled. This was the first commercially available general-purpose computer and was based on Williams and Kilburn’s work on the Manchester Baby and the Manchester Mark I.   

Next came Atlas – a joint development between The University of Manchester, Ferranti and Plessey – soon followed; it was one of the most powerful of its era, pioneering virtual memory and multiprocessing.

75 years of firsts

Manchester has been driving digital innovation ever since. Discover Manchester-made milestones:

1. Manchester Code (1949) – A data-encoding method still used in remote control consumer devices today.
2. First electronic music recording (1951) – Produced at Manchester with the BBC.
3. First computer game (1952) – Christopher Strachey draughts/checkers programme for the Manchester Mark 1.
4. First electronic literature (1952) – Strachey’s love-letter algorithm, a landmark in creative computing.
5. Virtual memory (1959) – Invented by Tom Kilburn, leading to the Atlas computer in 1962.
6. First UK computer science department (1964) – The University of Manchester opens the country’s first dedicated Department of Computer Science.
7. Alan Turing’s computational biology (1950s) – Groundbreaking research into morphogenesis. 
    

A community of changemakers

Our legacy of firsts continues today, reflected in the people who shape the future of computing and digital innovation:

• Steve Furber – Co-designer of the BBC Micro and the ARM processor and Professor here for over three decades; more than 230 billion ARM processors have been built worldwide.
• Pete Lomas – Alumni and Co-designer of the Raspberry Pi, which revolutionised computing education and innovation worldwide, putting affordable, programmable technology into the hands of millions of learners, makers, and entrepreneurs.
• Kim Libreri – Alumni and CTO of Epic Games, a global leader shaping the future of gaming through blockbuster titles like Fortnite
• Zahra Montazeri – lecturer in computer graphics, who’s research in rendering was used in The Mandalorian and Avatar: The Way of Water.
• , a Turing Fellow in the Department of Mathematics, whose benchmark software underpins much of the global supercomputing, making it possible for scientists, engineers, and governments to run large-scale simulations

Looking ahead: AI at Manchester

With more than 75 years of breakthroughs, The University of Manchester continues to push the boundaries of AI, from advancing core research to driving real-world impact. 

Our work today spans:

• Cutting-edge AI research in fields from healthcare to climate science.
• Industry collaborations accelerating innovation.
• Initiatives supporting inclusive economic growth.

Manchester is, and always has been, a powerhouse shaping the digital future for the UK and the world.

One tool, ‘Baler’, works with an autoencoder – a type of neural network trained to decrease the number of dimensions of input data, making it smaller. 

Caterina Doglioni, Professor of Particle Physics, explains: “There are multiple avenues to reduce the computing resources we use. One is reducing the amount of data to be stored through data compression.”  

The team are also measuring the energy usage of Baler and other approaches, to identify optimisations that could foster more energetically sustainable, data-driven scientific practices.  

Rosie Schiffmann, an undergraduate student in the research team, adds: “With Baler and data compression as an example, we’re giving researchers a way to track their computational ‘metabolism’ and make it more efficient. Green computing isn’t a futuristic vision; it’s actionable today if we rethink how we store and process data.”

The work in this project received funding from the European Union’s Horizon 2020 research and innovation programme and European Research Council (ERC) under Grant Agreements n. and .

Meet the researcher

The project is led by Caterina Doglioni, Professor of Particle Physics, together with supervisors James Smith (Postdoctoral Research Associate) and Michael Sparks (Senior Research Software Engineer). Within the University of Manchester team are PhD student Pratik Jawahar, Jack Goodsall and Rosie Schiffmann from the Physics & Astronomy internship program, Bradley Booth from DeepMind’s AI Fundamentals Summer Internship program, and Sakshi Kumar, a Google Summer of Code student, working with collaborators in the US, Sweden and Ukraine. 

• Email Prof Caterina Doglioni >>
• Email the Business Engagement Team >>

Read her papers

Most facial recognition models are trained once and then expected to work across every scenario. However, a model trained on one dataset often struggles when faced with new situations, and retraining from scratch is slow and inefficient. 

Dr Rahul Singh Maharjan and his team are tackling this challenge by developing a new approach: teaching AI to learn emotions incrementally. Instead of forgetting what it already knows, the system builds on past experiences whilst adapting to fresh emotional data. This makes it more resilient and better prepared for real-world human interaction. 

As Dr Maharjan explains: "For technology to truly integrate into our lives, it must understand our emotions. My goal is to help build AI that doesn’t just compute, but connects with us." 

The project she began works with AI to identify women experiencing menopause symptoms in Greater Manchester. It looks at how treatments of menopause vary by background, and aims to ensure that all women can access the support they need. 

Dr Woolley records insights from women and health professionals on what matters most in menopause experiences. Using AI, their insights are used to extract the most relevant information from big health datasets, like UK Biobank and the Greater Manchester data environment. This approach helps to ensure that women’s voices are directly shaping the research. 

And Dr Woolley believes that can make a real difference: 

“I was moved by women that told me about their experiences of unequal access to services, lack of information, misdiagnosis and inadequate treatment during the menopause. By combining women’s lived experiences with the power of AI and big data, my research will provide evidence that can drive change towards menopause care that is better informed and equitable for all.”

Meet the researcher

Dr Charlotte Woolley is an epidemiologist and Research Fellow for Manchester’s Healthier Futures Research Platform. Listed as an AI Visionary by the Department of Health and Social Care (DHSC) in 2025, for her pioneering work in women’s health and gender equity through artificial intelligence, Dr Woolley’s research is driven by her passion for women's health. She incorporates the real-life experiences of clinicians and women to guide the objectives of her work and interpret her findings. 

• Email Dr Charlotte Woolley>>
• Email the Business Engagement Team >>

Read her papers

As people increasingly search for connection in an often-isolated modern world, the line between technology and companionship is blurring. By examining what happens when those bonds with AI falter, Dr Cearns’ work sheds light not only on the ethics of human-machine intimacy, but also on the wider human search for belonging.

In her most recent project, she has used digital ethnography and interviews to examine how users emotionally invest in AI ‘soulmates’ – AI chatbots that become romantic partners to humans – and the grief that follows their malfunction or shutdown. This research is critical for highlighting new forms of kinship and ethical care in human-machine relationships.

Meet the researcher

Jennifer Cearns is Lecturer in AI Trust and Security, in the Department of Social Anthropology. Her research explores how people form emotional, romantic, and therapeutic relationships with AI, focusing on kinship, ethics, and cultural understandings of personhood.

• Email Dr Cearns >>
• Email the Business Engagement Team >>

Read her papers

That ethos has inspired a new project bringing fashion and technology together: designing custom-fit clothing for different body shapes.

By combining 3D body scanning, digital pattern cutting, and virtual fitting, the project delivers faster, better-fitting solutions that reduce fabric waste and promote inclusive, sustainable fashion.  

The process begins with 3D body scanning to capture accurate body measurements and shapes, which are then translated into digital patterns. These patterns are refined and tested through AI-enabled virtual fitting, allowing adjustments to be made without the need for physical samples.  

AI-powered tools within Clo3D further enhance this workflow: the AI pose generator creates realistic body postures for fit evaluation; the 3D garment simulation predicts fabric behaviour during movement; the AI-assisted range design automates size adjustments and style variations; and the Clo AI Studio accelerates ideation by generating design options. Together, these technologies integrate human creativity with AI-driven efficiency, ensuring precision, inclusivity, and sustainability throughout the design process. 

Led by Phumza Ntombovuyo Sokhetye, a PhD researcher in Textiles and Apparel, the work is transforming the frustrating trial-and-error processes for finding jeans, into custom designs that celebrate diversity. 

Building on the University’s strengths in sustainable innovation, Phumza describes the aim of the project as to create “perfectly fitting, eco-friendly clothing accessible to everyone, empowering all individuals to feel confident and included no matter their shape or size.” 

Pepper, who not only has human linguistic skills, but with recent developments in AI can now interact with people and even read emotions, is part of a project partnership between Age UK and Manchester’s Faculty of Science and Engineering. 

They believe that the ‘social robots’ they’re working on, can be used as companions to support health and care in older adults, as well as children with some disabilities. 

Sue Agar, Service Development Manager at Age UK Manchester  explains: “You can see the concern disappear quite quickly when they have the chance to speak with Pepper. It’s like a barrier comes down, and before long they treat Pepper like they would any other visitor.” 

The robots aren’t yet fully autonomous – Pepper is accompanied by staff and students from the University – but there’s a huge amount they can do independently. Pepper can communicate seamlessly with residents, understanding what they say and responding appropriately.   

And this isn’t the first robot visitor to Brunswick Village, previously a smaller model had been brought in to demonstrate a Tai Chi session.  

Leading the project is Professor Angelo Cangelosi, whose father suffered from dementia, and who sees a real potential in the future role this technology could play.  

He explains: “We live in a society with ever increasing needs for integrated health and social care solutions, to support healthy ageing. Social robots and AI can support such needs, within a human-centric approach putting people at the core of the development of trustworthy care solutions.” 

Though robots aren’t going to be a substitute for nurses and carers, Professor Cangelosi believes they can be used as a tool to support these professions. Potential roles could include monitoring illnesses in patients, helping people to access medications, or simply being a companion within the home. 

Sue Agar, Service Development Manager at Age UK Manchester adds: “There’s a huge amount of good the robots can do keeping people safe. The social intelligence is so important too, because loneliness is a real issue and companionship – being able to have conversation and interaction – makes such a difference.”  

With carers already under significant strain due to staff shortages, and with an ageing population very likely to increase in the years ahead, Professor Cangelosi and his team are working on robots that could play a vital role in reducing the growing pressure on the care sector.

[The University of Manchester has received a prestigious grant from the European Research Council to support this project, focussed on helping robots to understand more abstract concepts.] 

"AI has the potential to reduce workload in general practice, yet despite that potential, AI tools are not yet routinely used.” 

To explore current attitudes to the technology, Dr Brown and a team of researchers from The University of Manchester and Cambridge, ran a study around the use of AI in ‘eVisits’ – online consultations available to NHS patients. 

Known as ‘Patchs’, this AI uses Natural Language Processing and machine learning to analyse patient messages and understand decisions made by GPs. 

Participants identified seven opportunities for AI during their consultations, including sending patient requests to the most appropriate staff member and asking targeted follow-up questions to speed up the help they receive. 

Whilst the study’s lead author, Manchester’s Dr Moschogianis, says there were “concerns about the capacity of AI to deal with the complexity of primary care and fears of depersonalised service”, where it could be shown that the technology was supporting doctors and speeding up help, it was broadly welcomed by patients.   

With these positive results, the team feel that they’ve provided the first clear roadmap for developing AI tools that are both effective and trusted by patients. 

But spotting patterns in where and when these risks are highest isn’t easy. So, a team of researchers at Manchester led by Dr Zhonghua Zheng, have begun to design tools that help cities track these risks and adapt to growing climate and environmental challenges.  

Dr Zheng explains: “We urgently need tools that are not only accurate, but accessible and actionable. This project reflects my passion for using AI and open science to empower decision-makers, from local councils to the global research community.” 

By combining open data with a use of AI and detailed computer models, the team are creating more accurate tools that not only track and predict heat and air pollution in cities, but also evaluate the effectiveness of potential engineering solutions – helping leaders take action sooner, make better decisions, and build cleaner, healthier and more resilient urban futures.

As part of this project, Dr Zhao, Lecturer in Generative AI for Education at Manchester’s Institute of Education, is running a global survey to give educators and students a voice on how these standards continue to be shaped.  

Dr Zhao’s mission is to change a current situation that sees many AI policies written from the top down, with little input from practitioners and users.  

Working with UNESCO, she designed and led the survey, and is analysing the early results.  

The survey results will inform the design of her recently awarded ITL AI Fellowship at the University of Manchester. Drawing on insights from the global survey and guided by the UNESCO AI Competency Frameworks, she will develop a scalable programme to support staff and students in building AI competency.  

In collaboration with the University Library, the programme will create a non-judgemental space that encourages deep reflection on their use of AI and its outcomes will be shared with JISC to support collaboration on AI competency training across the wider higher education sector. 

Reflecting on this work, Dr Zhao explains: “In the age of AI, we face both new opportunities and complex challenges. To navigate this, we need the right skills and a responsible, ethical relationship with AI in society. This project enables me to gather global insights from teachers and students, supporting UNESCO in shaping AI guidelines that can guide universities around the world.” 

Manchester’s Dr Christopher Bowden and Dr Tim Foster have set out to answer this question, applying machine learning algorithms and high-resolution satellite imagery to identify where farmers in Ghana use irrigation – revealing where communities have expanded irrigation systems or where improved water access could transform crop productivity.  

This data-driven approach ensures irrigation services reach the farmers in greatest need and represents a strong example of blending research with impact. Dr Foster is pleased with the real-world effect the project has already had: 

 “We can now rapidly map and monitor where and when farmers are adopting irrigation in Ghana and other African countries. We use these maps to help governments, development agencies, NGO’s and the private sector to better design and target irrigation projects, to improve food security and help reduce rural poverty.” 

Yet Dr Andrea Lagna, an expert in Information Systems, is challenging the assumption that we have to choose between AI development and a sustainable planet. 

Dr Lagna applies prospective theorising to his work: rather than limiting his research to analysing past results, he undertakes an imaginative, value-driven, and evidence-based exploration of how business organisations can balance AI innovation with environmental stewardship.  

Through this approach, he imagines a world where technological innovation and the responsible management of our resources go hand-in-hand. Solutions might be found within the use of more energy-efficient tools and by fostering multi-stakeholder governance, where diverse groups are included in decision-making, leading to more balanced outcomes.  

Dr Lagna champions the idea that organisations can transform environmental responsibility into a source of competitive advantage, in part because they must. He explains: “This alignment is the most critical strategic objective for business organisations in our time of climate crisis.”

At The University of Manchester, researchers are pioneering a new approach to make this possible. Known as Credal Bayesian Deep Learning (CBDL), it allows AI to recognise and communicate how confident – or uncertain – it is about a decision. Unlike traditional neural network systems, which often act as if they’re always sure, CBDL can separate situations where more data could improve accuracy, from those where uncertainty will always remain. 

CBDL does this by training a set of neural networks that work together, producing not just a single answer but a range of possible outcomes within probability bands. This gives engineers and doctors a clearer picture of what an AI system really knows, and where caution is needed. 

As Manchester researcher Dr Michele Caprio explains: “Knowing what a model does not know is crucial for safety-critical AI. That transparency is the foundation for certifiable autonomy in cars, insulin pumps, and beyond.”

In Ethiopia, TFGBV has become a serious challenge, yet little quantitative evidence has existed to measure its scale or provide solutions. Aiming to fill that gap, a Manchester team led by Dr Riza Batista-Navarro, in collaboration with the Centre for Information Resilience, carried out the ADAGE project to highlight the scale and nature of gendered hate speech online. 

Natural language processing (NLP) 

To carry out the research, CIR developed a lexicon of more than 2,000 inflammatory terms across four languages – Amharic, Afaan Oromo, Tigrigna and English.  

Then, by combining expertise in computational linguistics, NLP and the Ethiopian online context, Dr Riza Batista-Navarro’s team developed a framework for identifying hate-containing posts on social media, while factoring in dimensions such as the target, type and nature of hate speech. 

This approach enabled the analysis of millions of social media posts, of which more than 7k were examined in detail. The analysis led to key findings: (a) that – different to Ethiopian men – Ethiopian women receive substantial hate speech in the form of mockery, irony and gender stereotypes that imply inferiority; and (b) the risk of women being targeted by online hate speech is compounded by other protected characteristics such as ethnicity. Working closely with Ethiopian experts, the team ensured cultural and linguistic accuracy, producing the first large-scale labelled dataset of its kind. 

Data to inform action 

The findings show that women and girls face distinct forms of online abuse compared to men and boys. Gendered insults, stereotypes, and mockery are commonplace, often minimised or dismissed as less harmful than threats or aggressive language. Yet these forms of abuse reinforce harmful gender norms and contribute to the silencing of women in public life. Intersectional abuse, where gender combines with ethnicity or religion, was also prevalent, particularly during times of conflict. 

Addressing TFGBV is vital to ensuring women and girls can participate safely and meaningfully in public life.  

The project has already led to a report and a set of 34 recommendations across seven policy areas, designed to guide government, civil society and tech companies. Together, they offer practical recommendations towards a safer online environment – and greater gender equality. These recommendations include: targeted, platform-specific responses; greater public education on hate speech; and stronger action from governments, civil society organisations, and social media companies are required. 

By strengthening the evidence base and providing practical recommendations, the ADAGE project has helped support safer online spaces for women and girls in digital and public life.