Major depressive disorder, dementia, anxiety disorders, and substance abuse affect a substantial part of the European older population. Over 70% of Europeans reside in cities, and this percentage will increase in the next decades. Urbanization and ageing have enormous implications for public mental health. Cities pose major challenges for older citizens, but also offer opportunities for the design of policies, clinical and public health interventions that promote mental health. The overall aim of the MINDMAP project is to identify the opportunities offered by the urban environment for the promotion of mental wellbeing and cognitive function of older individuals in Europe. The project will advance understanding by bringing together longitudinal studies across cities in Europe, the US and Canada to unravel the causal pathways and multi-level interactions between the urban environment and the social, behavioural, psychosocial and biological determinants of mental health and cognitive function in older adults. Specifically, the project will (a) assess the impact of the urban environment on the mental wellbeing and disorders associated with ageing, and estimate the extent to which exposure to specific urban environmental factors and policies explain differences in ageing-related mental and cognitive disorders both within as well as between European cities, (b) assess the causal pathways and interactions between the urban environment and the individual determinants of mental health and cognitive ageing in older adults, (c) use agent-based modelling to simulate the effect of urban environmental, prevention and care policies on the trajectories of mental health and cognitive ageing across cities in Europe. Knowledge will significantly contribute to future-proof preventive strategies in urban settings favouring the mental dimension of healthy ageing, the reduction of the negative impact of mental disorders on co-morbidities, and maintaining cognitive ability in old age.

Deep learning (DL) is the key technique in modern artificial intelligence (AI), which has provided state-of-the-art accuracy on many machine-learning based applications. Today, although most of the computational loads of DL systems are still spent running neural networks in data centres, the ubiquity of smartphones, and the upcoming availability of self-contained wearable devices for augmented reality (AR), virtual reality (VR) and autonomous robot systems are placing heavy demands on DL-inference hardware with high energy and computing efficiencies along with rapid development of DL techniques. Recently, we have witnessed a distinct evolution in the types of DL architecture, with more sophisticated network architectures proposed to improve edge AI inference. This includes dynamic network architectures that change with each new input in a data-dependent way, where inputs and internal states are not fixed. Such new architectural concepts in DL are likely to affect the type of hardware architectures that will be required to deliver such capabilities in the future. This project precisely addresses this challenge and proposes to design a flexible hardware architecture that enables adaptive support for a variety of DL algorithms on embedded devices. Primarily, to produce lower cost, lower power and higher processing efficiency DL-inference hardware that can be configured adaptably for dedicated application specifications and operating environments, this will require radical innovation in the optimisation of both the software and the hardware of current DL techniques. This work aims to perform fundamental research, development and practical demonstrator to enable general support for a variety of DL techniques on embedded edge devices with limited resource and latency budgets. Primarily, this requires radical innovation on the current DL architectures in terms of computing architecture, memory hierarchy and resource utilisation, as well as system latency and throughput: it is particularly important for the modern DL systems that the inference processes are dynamic, such as, the DL inference maybe input-dependent and resource-dependent. The proposal therefore seeks the following three thrusts: First, to build upon the existing work of the PI in optimising machine-learning models for resource-constrained embedded devices, towards achieving the goal that the network model could be dynamically optimised as needed through hardware-aware approximation techniques. Second, with newly-developed adaptive compute acceleration technology in programmable memory hierarchy and adaptive processing hardware, to seek a new ambitious direction to develop a set of context-aware hardware architectures to work closely with the approximation algorithms that can fully utilise the true hardware capabilities. Unlike traditional optimisation techniques for DL hardware inference engines, the proposed work will explore both software and hardware programmability of adaptive compute acceleration technology, in order to maximise the optimisation results for the target application scenarios. Third, this project will work closely with our industry and project partners to produce a practical demonstrator to showcase the effectiveness of the proposed DL framework versus traditional approaches, particularly, evaluating the effectiveness of the framework in real-world mission-critical applications.

Changes to our society are being driven by both long-term social and economic trends, and the impacts of recession and austerity. Five social trends drive the new research programme of the Research Centre on Micro-Social Change (MiSoC): i) new features of the job market that are changing the fortunes of different groups of people; ii) changing family set-up and relationships; iii) reform of and cut-backs in the provision of housing, education, health, and benefits; iv) breaking down of social and political beliefs, and increasing ethnic and religious diversity; v) changes in our values to be more accepting of personal freedom and more tolerant of inequality. Of course, modern societies are always changing, but the next decade poses new challenges. Recession, austerity and the patchy nature of the recovery mean things looks bleak for many. Ties between family, friends and neighbours, weakening as the UK grew richer and as individuals became more mobile, have been put under further stress by hard times. Our new research programme aims to point to ways in which society can continue to integrate people with diverse backgrounds, preferences and abilities. The research will be led by a team of experts at the Institute for Social and Economic Research (ISER) at the University of Essex, with collaborators across the UK and in other countries, and will be directed jointly by Mike Brewer, Professor of Economics and David Voas, Professor of Population Studies. Our work covers three main areas: The first area examines how individuals and families are affected by and react to changes in their life circumstances, including shocks to their health, disability, income and living arrangements. Our researchers will pay special attention to the way that new welfare systems, such as changes to benefits, protect households. We will be making a major contribution to important debates on poverty by advancing new ways of measuring poverty, and with new evidence on the dynamics of poverty. The second investigates how new members of society - children, young people and new migrants -develop and are integrated into it. We will analyse how parents, school, peers and society interact to influence the development of children's mental, social and physical skills, and the long-term consequences of childhood disadvantage. We will also look at how some people get more out of gaining a university degree than others. We will provide new evidence on the integration of ethnic minorities, and how this varies across individuals. We will look at the experience of new migrants, and how characteristics and behaviours are passed between generations in migrant families. The third area of research investigates how values, attitudes, expectations, tastes or preferences and identity are formed, and how they are linked to our education, employment and family set-up. A better understanding of this will help policy-makers come up with the best policies to help more people live successful, happy lives. How we research the important issues facing society today is just as important as the research itself, so our integrated programme of methodological work will help researchers to better examine the impact of specific policies, and to advise on new ways to handle the sometimes incomplete information which comes from survey data they are using in their research. We expect this programme of research to benefit a wide range of organisations involved in policy debates, policy design and practice, in a range of domains, located in the UK and other countries, and provide evidence informing key policy choices, such as the balance between intervening late or early in children's lives, the role of family and wider society in an individual's development, the choice between universal or targeted support or safety nets for the vulnerable, and the relative roles of values, expectations and preferences versus structure in determining how we act.

Doctoral Training Partnerships: a range of postgraduate training is funded by the Research Councils. For information on current funding routes, see the common terminology at https://www.ukri.org/apply-for-funding/how-we-fund-studentships/. Training grants may be to one organisation or to a consortia of research organisations. This portal will show the lead organisation only.

The dispensing of hearing aids has not kept pace with the dramatic increase in technological sophistication of the aids themselves. Current diagnostic tests are simple, the principles for choosing a particular aid are rudimentary, the evaluation of hearing after dispensing is inadequate and, not surprisingly, levels of patient satisfaction are much lower than for, for example, spectacles.This project will develop a new approach to the problem by adapting a computer model of hearing to represent the particular deficit of an individual patient. The model (or 'hearing dummy'') can then be used to evaluate objectively, the potential benefit of different hearing aid designs and indicate a 'best-buy' prescription.The aim is to produce a comprehensive system from assessment through to dispensing advice that is viable in a clinical context given the normal constraints of audiological practice. Within this general aim, a particular effort is required in three areas, patient assessment, adapting the computer model to represent the hearing of the patient and using the model to generate dispensing advice. Patient assessment requires a range of auditory tests that can be conducted quickly which cover a range of different auditory functions. The test battery will build on recently-developed assessment protocols and will cover functions not normally assessed in audiology clinics at present but are necessary to define the underlying pathology. They will include the measurement of filter bandwidths, compression, temporal integration, distortion-product oto-acoustic emissions and averaged brain response in addition to absolute thresholds. These tests will need to be faster and easier to administer than existing laboratory techniques. In this respect we will benefit from the support of the local audiology clinic and a private dispensing agency where the practicalities of assessment with patients are their primary concern.An existing computer model of normal peripheral hearing will be used in this project. It will be adapted using data collected in the assessment stage to represent the hearing of that individual patient. The accuracy of the 'hearing dummy' model will be assessed by testing it using the same audiometric procedures used with the patient. The model must give the same outcome as the patient tests to be regarded as accurate. While this may appear to be a tall order, the results of 20 years of research and development of this particular model offer reassurance that it is achievable on a routine clinical basis.A hearing aid transforms the ambient acoustic signal into a form thought to be more useful to the patient. Some transforms will be more successful than others in restoring the response of the auditory periphery to a more normal pattern. The project will measure the response of the model to different signal transforms representing different hearing aid designs. These outputs will be compared to the response of a model of a healthy ear to the original sound. The best aid is forecast as the one that restores the output to a pattern closest to normal.A good hearing aid is one that helps the patient function normally in everyday situations. The greatest challenge is to hear speech details against a confusing acoustic background. The project will focus on this as the ultimate test of a beneficial aid. In this we shall benefit from a parallel collaborative project with Sheffield University that is using our auditory computer model to develop and evaluate automatic recognition of speech sounds in noisy backgrounds.