Machine learning offers great promise in helping us solve problems by automatically learning solutions from data, without us having to specify all details of the solution as in earlier computational approaches. However, we still need to tell machine learning systems what problems we want them to solve, and this is currently undertaken by specifying desired outcomes and designing objective functions and rewards. Formulating the rewards for a new problem is not easy for us as humans, and is particularly difficult when we only partially know the goal, as is the case at the beginning of scientific research. In this programme we develop ways for machine learning systems to help humans to steer them in the process of collecting more information by designing experiments, interpreting what the results mean, and deciding what to measure next, to finally reach a conclusion and a trustworthy solution to the problem. The machine learning techniques will be developed first for three practically important problems and then generalized to be broadly applicable. The first is diagnosis and treatment decision making in personalized medicine, the second steering of scientific experiments in synthetic biology and drug design, and the third design and use of digital twins in designing physical systems and processes. An AI centre of excellence will be established at the University of Manchester, in collaboration with the Turing Institute and a number of partners from the industry and healthcare sector, and with strong connections to the networks of best national and international AI researchers.

An increasing number of people live with long term physical and mental health conditions, such as diabetes, heart disease or depression. Many of these people find that their symptoms fluctuate in severity over time, including periods of relative calm and episodes during which symptoms become much worse. However, patients with long term conditions typically see their doctor during pre-arranged visits at fixed intervals, rather than on the basis of their current symptoms. For instance, people with chronic kidney disease commonly have appointments every 3 months. These visits are often felt unnecessary during stable periods, during which patients could probably manage well by themselves, but irregular enough to spot worsening symptoms early enough and prevent more severe episodes of illness - what we call 'fall back episodes'. We propose to develop a set of software tools for smartphones and tablets, called the "Wearable Clinic". This will help patients with long term conditions, together with their carers and doctors, to better manage their health in daily life, respond more quickly to changes in symptoms and prevent fall back episodes. This could prevent unplanned admissions to hospital, which are not only distressing and disruptive for patients and their families, but expensive for the NHS. Furthermore, it could make it easier to integrate care for patients with multiple long term conditions (e.g. both diabetes and chronic kidney disease), who are often treated by different doctors, at different places, and at different times. For patients, using the Wearable Clinic starts with measuring symptoms in daily life using wearables. These data are then automatically combined with data held in NHS records on their diagnoses, lab results, and treatments in order to predict the likely future course of symptoms, and whether there is a risk of a fall back episode. Finally, the software will propose a modifiable care plan that takes account of the patient's range of existing conditions, current and predicted health status, availability of local care resources, and the patient's own preferences. Where it is possible and safe to do so, care plans will remove clinically unnecessary and unwanted appointments, saving time and money for both the patient and the NHS. To achieve this vision, we propose to apply data science techniques to analyse data collected from a) medical records and b) wristband wearables and smartphone technologies ('wearables') worn by patients with long term conditions. While the Wearable Clinic concept could potentially be useful for managing a range of long term conditions, we will first test it out in two different conditions, where symptoms are known to fluctuate over time: schizophrenia and chronic kidney disease. Statistical techniques will be applied to see if data collected from patients using wearables can be used to a) predict changes in symptoms and b) produce tailored care plans for individual patients. We will trial methods that collect and use data in ways that take into account individual risk factors (e.g. age, ethnicity) and conserve the battery life of devices. While the project primarily aims to develop new computer algorithms, statistical models and computer software, we will trial the technical aspects of the Wearable Clinic with a small number of healthy volunteers, people with schizophrenia and people with chronic kidney disease. We will also investigate costs, benefits, and potential risks of the Wearable Clinic in its earliest stages of development and, where necessary and feasible, integrate solutions during the lifetime of the project. A series of workshops open to the public will be held to explore cross-cutting issues such as trustworthy data use and privacy. This will pave the way for future studies and maximise the chances that the Wearable Clinic actually makes it into practice - thus improving the quality of care for patients with long term conditions.

The World Health Organisation says that there are about 100 million people globally who need prosthetic or orthotic (P&O) services and as populations age, more than two billion people are expected to require health-related assistive devices by 2030. In the UK the Disabled Living Foundation estimates that 6.5 million people live with mobility disablement, with many reliant on P&O services, including an estimated two million orthotic users. In parts of the developing world the aftermath of conflict, such as land mines, and greater rates of traumatic injuries from accidents, means there is a growing need for prosthetics and orthotics for younger people living in poor social and economic circumstances. Often they need P&O devices to stay at work and sustain their families. Poor devices, services and access to these contravene their basic human rights. In the context of this need, we want to establish the EPSRC Centre for Doctoral Training in P&O. This will address the national, and global, shortage of suitably skilled engineers and scientists to become future innovators in P&O technologies. Current academia, industry and care centres have limited researchers, and research activity has lagged behind rapid technology advancements. The Centre will support a minimum of 58 doctoral students whose studies will enable them to become leaders of the future. The Centre will bring together the only two P&O undergraduate education facilities in the UK (Salford and Strathclyde) with P&O research centres of excellence at Imperial College and the University of Southampton. Our vision is for the Centre to become the national and global leader in P&O research training, and the translation of research into innovation that impacts on the lives of people each day, in developed and developing countries. The Centre will work to support training for students from low and middle-income countries (LMIC). Our students will be immersed in industry and real-world experiences which will equip them to lead the P&O sector across technology, social or economic contexts. Our aims are to: 1. Develop a new model of P&O research training and translation of research into innovation. In addition to the doctoral training, this will result in Master's programmes operating across Institutions. 2. Produce ambitious PhD research projects that will be grounded in real-world challenges, but at the cutting-edge of new biomedical science and technologies. 3. Produce a significant impact on the UK P&O industry sector by leading innovation. 4. Have an international impact by attracting an increasing number of CDT students from overseas. 5. Establish a P&O student society which will have matured into a lasting doctoral community with international reach. 6. To have a significant impact on the training of doctoral candidates from LMIC. 7. Attract additional external funding for P&O research. Creating a new generation of P&O research leaders will, over time, have a significant economic, societal and health impact. For users, it will mean access to improved generations of assistive devices which will match the users' needs resulting in a better quality of life. Clinical services will benefit from improved service data, superior products and improved user outcomes. For industry, it will open up new market opportunities, nationally and globally. For the students themselves, they will have access to careers that have a real purpose, enabling them and their future teams to make a difference in the lives of people with disabilities.