This project will propose an urban grammar to describe urban form and will develop artificial intelligence (AI) techniques to learn such a grammar from satellite imagery. Urban form has critical implications for economic productivity, social (in)equality, and the sustainability of both local finances and the environment. Yet, current approaches to measuring the morphology of cities are fragmented and coarse, impeding their appropriate use in decision making and planning. This project will aim to: 1) conceptualise an urban grammar to describe urban form as a combination of "spatial signatures", computable classes describing a unique spatial pattern of urban development (e.g. "fragmented low density", "compact organic", "regular dense"); 2) develop a data-driven typology of spatial signatures as building blocks; 3) create AI techniques that can learn signatures from satellite imagery; and 4) build a computable urban grammar of the UK from high-resolution trajectories of spatial signatures that helps us understand its future evolution. This project proposes to make the conceptual urban grammar computable by leveraging satellite data sources and state-of-the-art machine learning and AI techniques. Satellite technology is undergoing a revolution that is making more and better data available to study societal challenges. However, the potential of satellite data can only be unlocked through the application of refined machine learning and AI algorithms. In this context, we will combine geodemographics, deep learning, transfer learning, sequence analysis, and recurrent neural networks. These approaches expand and complement traditional techniques used in the social sciences by allowing to extract insight from highly unstructured data such as images. In doing so, the methodological aspect of the project will develop methods that will set the foundations of other applications in the social sciences. The framework of the project unfolds in four main stages, or work packages (WPs): 1) Data acquisition - two large sets of data will be brought together and spatially aligned in a consistent database: attributes of urban form, and satellite imagery. 2) Development of a typology of spatial signatures - Using the urban form attributes, geodemographics will be used to build a typology of spatial signatures for the UK at high spatial resolution. 3) Satellite imagery + AI - The typology will be used to train deep learning and transfer learning algorithms to identify spatial signatures automatically and in a scalable way from medium resolution satellite imagery, which will allow us to back cast this approach to imagery from the last three decades. 4) Trajectory analysis - Using sequences of spatial signatures generated in the previous package, we will use machine learning to identify an urban grammar by studying the evolution of urban form in the UK over the last three decades. Academic outputs include journal articles, open source software, and open data products in an effort to reach as wide of an academic audience as possible, and to diversify the delivery channel so that outputs provide value in a range of contexts. The impact strategy is structured around two main areas: establishing constant communication with stakeholders through bi-directional dissemination; and data insights broadcast, which will ensure the data and evidence generated reach their intended users.

Civil infrastructure is the key to unlocking net zero. To achieve the ambitious UK targets of net zero by 2050, we require innovative approaches to design, construction, and operation that prioritise energy efficiency, renewable resources, and low-carbon materials. Meeting net zero carbon emissions will require not only significant investment and planning, but also a radical shift in how we approach the design and management of our civil infrastructure. Reliable low carbon infrastructure sector solutions that meet real user needs are essential to ensure a smooth and safe transition to a net zero future. To address these challenges, the UK must develop highly skilled infrastructure professionals who can champion this urgent, complex, interconnected and cross-disciplinary transition to net zero infrastructure. This EPSRC Centre for Doctoral Training in Future Infrastructure and Built Environment: Unlocking Net Zero (FIBE3 CDT) aims to lead this transformation by co-developing and co-delivering an inspirational doctoral training programme with industry partners. FIBE3 will focus on meeting the user needs of the construction and infrastructure sector in its pursuit of net zero. Our goal is to equip emerging talents from diverse academic and social backgrounds with the skills, knowledge and qualities to engineer the infrastructure needed to unlock net zero, including technological, environmental, economic, social and demographic challenges. Achievable outcomes will include a dynamic roadmap for the infrastructure that unlocks net zero, cohort-based doctoral student training with immersive industry experience, a CDT which is firmly embedded within existing net zero research initiatives, and expanded networks and outward-facing education. These outcomes will be centred around four thematic enablers: (1) existing and disruptive/new technologies, (2) radical circularity and whole life approach, (3) AI-driven digitalisation and data, and (4) risk-based systems thinking and connectivity. FIBE3 doctoral students will be trained to unlock net zero by evolving the MRes year to include intimate industry engagement through the novel introduction of a fourth dimension to our successful 'T-shaped' training model and designing the PhD with regular outward-facing deliverables. We have leveraged industry-borne ideas to align theory and practice, streamline business and research needs, and provide both academic-led and industry-led training activities. Cohort-based training in technical, commercial, transferable and personal skills will be provided for our graduates to become skilled professionals and leaders in delivering net zero infrastructure. FIBE3's alignment with real industry needs is backed by a 31 strong consortium, including owners, consultants, contractors, technology providers and knowledge transfer partners, who actively seek engagement for solutions and will support the CDT with substantial cash (£2.56M) and in-kind (£8.88M) contributions. At Cambridge, the FIBE3 CDT will be embedded within an inspirational research and training environment, a culture of academic excellence and within a department with strategic cross-cutting research themes that have net zero ambitions at their core. This is exemplified by Cambridge's portfolio of over £60M current aligned research grant funding and our internationally renowned centres and initiatives including the Digital Roads of the Future Initiative, the Centre for Smart Infrastructure and Construction, Cambridge Zero and Cambridge Centres for Climate Repair and Carbon Credits, as well as our strong partnerships with UK universities and leading academic centres across the globe. Our proposed vision, training structure and deliverables are exciting and challenging; we are confident that we have the right team to deliver a highly successful FIBE3 CDT and to continue to develop outstanding PhD graduates who will be net zero infrastructure champions of the future.

The Places of Poetry will create a distinctive digital map of England and Wales, onto which crowd-sourced poems of place, heritage and identity will be pinned in the course of a public campaign in the late-spring and summer of 2019. The project aims to prompt reflection on national and cultural identities in England and Wales, celebrating the diversity, heritage and personalities of place. The project thus combines a model from the past - the early seventeenth-century epic of national description, Michael Drayton's Poly-Olbion - with a commitment to the value of creative practice in the present day. Poly-Olbion has been the subject of an editorial and critical project led by the PI on this application, Andrew McRae. That project also involved some public-facing work, supported further by the Heritage Lottery Fund, which helped to demonstrate the power today of both Drayton's work and the distinctive county maps published with the poem, by the engraver William Hole. Meanwhile, the poet Paul Farley (Co-I on this project), has spent a number of years working on a twenty-first-century reconceptualization of Poly-Olbion (due for publication with Faber in 2019). This led to the collaboration at the heart of 'The Places of Poetry'. The new map of England and Wales will draw heavily upon the iconography of Hole's original works, but will be amended as necessary and lightly updated. Since Drayton and Hole only covered England and Wales, The Places of Poetry will also limit itself to these two British nations. The map will be overlaid upon Ordnance Survey data, with functions enabling users not only to zoom in and out, but also to slide between the two maps. Users will be encouraged to pin poems to particular places. The map will be pre-populated with a selection of historic (out of copyright) pieces; however, the project's central aim is to generate original work. The project website, with the map as its central and defining image, will aim to introduce users to the poetry of place, heritage and identity, and will provide materials designed to coach them through their own compositions. After a period of preparatory work, The Places of Poetry will be promoted through the course of an intensive campaign in the summer of 2019. This will include: a launch event, BBC radio programmes, a national writing-workshops tour, a short film, and social media activity. The project involves strategic collaborations with The Poetry Society and The Ordnance Survey. Although the project is technically freestanding, a positive response to the present application will trigger further applications (already in advanced states) to the Heritage Lottery Fund and the Arts Council, for a parallel package of work that would significantly enhance the reach and value of the project. This will be centred on activities with major regional arts and heritage partners, and involving a programme of poets-in-residence. It will focus on particular user-communities (e.g. age, location, background), and different kinds of heritage (e.g. pre-historic, Roman, agricultural, industrial, religious, natural). The website will remain open for contributions for ten weeks. Towards the end of this period, the PI will write a reflective article about the project. As a legacy, there may be opportunities for commercial development, while the site will remain in existence after the project closes, as an archive of the poetry of place, heritage and identity in the summer of 2019.

The project aims to create a prototype multi-sensor device, and undertake fundamental enabling research, for the location of underground utilities by combining novel ground penetrating radar, acoustics and low frequency active and passive electromagnetic field (termed quasi-static field) approaches. The multi-sensor device is to employ simultaneously surface-down and in-pipe capabilities in an attempt to achieve the heretofore impossible aim of detecting every utility without local proving excavations. For example, in the case of ground penetrating radar (GPR), which has a severely limited penetration depth in saturated clay soils when deployed traditionally from the surface, locating the GPR transmitter within a deeply-buried pipe (e.g. a sewer) while the receiver is deployed on the surface has the advantage that the signal only needs to travel through the soil one way, thereby overcoming the severe signal attenuation and depth estimation problems of the traditional surface-down technique (which relies on two-way travel through complex surface structures as well as the soil). The quasi-static field solutions employ both the 50Hz leakage current from high voltage cables as well as the earth's electromagnetic field to illuminate the underground infrastructure. The MTU feasibility study showed that these technologies have considerable potential, especially in detecting difficult-to-find pot-ended cables, optical fibre cables, service connections and other shallow, small diameter services. The third essential technology in the multi-sensor device is acoustics, which works best in saturated clays where GPR is traditionally problematic. Acoustic technology can be deployed to locate services that have traditionally been difficult to discern (such as plastic pipes) by feeding a weak acoustic signal into the pipe wall or its contents from a remote location. The combination of these technologies, together with intelligent data fusion that optimises the combined output, in a multi-sensor device is entirely novel and aims to achieve a 100% location success rate without disturbing the ground (heretofore an impossible task and the 'holy grail' internationally).The above technologies are augmented by detailed research into models of signal transmission and attenuation in soils to enable the technologies to be intelligently attuned to different ground conditions, thereby producing a step-change improvement in the results. These findings will be combined with existing shallow surface soil and made ground 3D maps via collaboration with the British Geological Society (BGS) to prove the concept of creating UK-wide geophysical property maps for the different technologies. This would allow the users of the device to make educated choices of the most suitable operating parameters for the specific ground conditions in any location, as well as providing essential parameters for interpretation of the resulting data and removing uncertainties inherent in the locating accuracy of such technologies. Finally, we will also explore knowledge-guided interpretation, using information obtained from integrated utility databases being generated in the DTI(BERR)-funded project VISTA.

On a daily basis huge amounts of geospatial data and information that record location is created across a wide range of environmental, engineered and social systems. Globally approximately 2 quintillion bytes of data is generated daily which is location based. The economic benefits of geospatial data and information have been widely recognised, with the global geospatial industry predicted to be worth $500bn by 2020. In the UK the potential benefits of 'opening' up geospatial data is estimated by the government to be worth an additional £11bn annually to the economy and led to the announcement of a £80m Geospatial Commission. However, if the full economic benefits of the geospatial data revolution are to be realised, a new generation of geospatial engineers, scientists and practitioners are required who have the knowledge, technical skills and innovation to transform our understanding of the ever increasingly complex world we inhabit, to deliver highly paid jobs and economic prosperity, coupled with benefits to society. To seize this opportunity, the Centre for Doctoral Training in Geospatial Systems will deliver technically skilled doctoral graduates equipped with an industry focus, to work across a diverse range of applications including infrastructure systems, smart cities, urban-infrastructure resilience, energy systems, spatial mobility, structural monitoring, spatial planning, public health and social inclusion. Doctoral graduates will be trained in five core integrated geospatial themes: Spatial data capture and interpretation: modern spatial data capture and monitoring approaches, including Earth observation satellite image data, UAVs and drone data, and spatial sensor networks; spatial data informs us on the current status and changes taking place in different environments (e.g., river catchments and cities). Statistical and mathematical methods: innovative mathematical approaches and statistical techniques, such as predictive analytics, required to analyse and interpret huge volumes of geospatial data; these allow us to recognise and quantify within large volumes of data important locations and relationships. Big Data spatial analytics: cutting edge computational skills required for geospatial data analysis and modelling, including databases, cloud computing, pattern recognition and machine learning; modern computing approaches are the only way that vast volumes of location data can be analysed. Spatial modelling and simulation: to design and implement geospatial simulation models for predictive purposes; predictive spatial models allow us to understand where and when investment, interventions and actions are required in the future. Visualisation and decision support: will train students in modern methods of spatial data visualisation such as virtual and augmented reality, and develop the skills on how to deliver and present the outputs of geospatial data analysis and modelling; skills required to ensure that objective decisions and choices are made using geospatial data and information. The advanced training received by students will be employed within interdisciplinary PhD research projects co-designed with 40 partners ranging from government agencies, international engineering consultants, infrastructure operators and utility companies, and geospatial technology companies; organisations that are ideally positioned to leverage of the Big Data, Cloud Computing, Artificial Intelligence and Internet of Things (IoT) technologies that are predicted to be the key to "accelerating geospatial industry growth" into the future. Throughout their training and research, students will benefit from cohort-based activities focused on group-working and industry interaction around innovation and entrepreneurship to ensure that our outstanding researchers are able to deliver innovation for economic prosperity across the spectrum of the geospatial industry and applied user sectors.