RIBA Think Piece SeriesDigital planningIdeas to make it happen

Planning (in) thedigitalised future

By Peter Stewart

Today, it feels as if we are still in the digital Stone Age when it comes to the planning process.

If you’ve been excited by the digital world created in Alfonso Cuaron’s Gravity, then looking up a planning application on a local authority website is likely to bring you back down to earth with even more of a bump than Sandra Bullock’s landing. What are the prospects for better and more sophisticated digitisation of the planning system?

It’s not hard to imagine amazing possibilities for spatial planning in a digitised world, given the continuing exponential growth of computing power and capacity that we can expect (even if Moore’s Law now turns out to have been more of a guideline – a bit like planning policies). The kind of digital imagery we are used to seeing on Time Team, with successive phases of building on an archaeological site reconstructed in ‘fast forward’ fly-throughs, could be used for future project proposals and made available for consultees to review on a local authority website. Or a dynamic imaging app could allow you to hold up your iPad in front of you on site and view a new scheme overlaid on reality, as it would appear from that viewpoint. Increasingly detailed digital city models already exist, and with more detailed data, greater computing power and better applications, the possibilities for inserting schemes in a digital world are exciting.

But today, it feels as if we are still in the digital Stone Age when it comes to the planning process. The applications suggested above wouldn’t need any technology we don’t have already (and probably exist already in some form) - but they are not likely to become standard practice soon. The reality is that digitisation of the planning system is in its infancy – and for the most part it is in the hands of local authorities, who are generally not at the bleeding edge of technology. The presentation of planning applications on a local authority website is typically poor: dumb search functions with no fuzzy logic, so you can’t find a site in the first place; bad indexing of documents if you do find them; and documents scanned at poor resolution, or in such large files that they are broken into dozens of parts for downloading.

In the spirit of learning to walk before you can run, I suggest we need to think about the digital near future a bit harder, and worry about the more distant future when the present system is working properly. What would be nice this year would be:

•Nationalstandardsforpresentingdata,sothatasearchonanylocalauthoritywebsitelooksthesame (or better, is a national resource) •StandardsofsearchfunctionsthatmatchthoseofsayGoogle–notmuchtoask,butthepresent reality is a long way from that; and all data geo-located on maps. •Clear,userfriendlypresentationofplanningdocuments,indigitaloriginalsratherthanscans,viewable online without needing to download, and all suitable for a lay person with a home PC.

All that would be a good start, but would appear to be some way off.

What about further ahead? Even as the physical reality of building proposals can be presented in more and more sophisticated ways through computer modelling, will this bring about better planning? It’s hard to see why one should expect that. The many problems of the UK planning system are not mainly to do with lack of access to data.

Digital exclusion, too, should be a major concern in a system that is supposed to be democratically accountable. Your 80 year old mother might want to say something about the Wetherspoon planned to open on her doorstep (mine did), but the average council website will not make it easy for her, even if she did use the internet.

In an optimistic version of the digital future, planning authorities will be much more readily able to receive data as well as to disseminate it. In that case, will voters still want their councillors deciding what will happen – why not decision-making by popular vote? Compared with a digital city model, the system that would allow citizens to vote online on planning applications and strategies would be pretty straightforward. But there is little appetite anywhere - least of all with the politicians who would have to give up the power they enjoy - for rule by plebi-scite rather than by representative government.

That might lead you to wonder what the point would be in providing citizens with increasingly sophisticated data concerning things they are not being asked to decide on in any case.

By the time we are ready to move to a more sophisticated level of digital planning, there will be hardware and software as yet undreamt of, so let’s worry about that when the time comes. A system that allows you access to the data you seek without a significant rise in blood pressure would be good for now. AuthorPeter Stewart, a chartered architect, is the principal of Peter Stewart Consultancy, a practice which provides expert advice on architecture, urban design and the historic environment, specialising in high profile projects and sensitive sites. He was the Director of the design review programme at CABE from 1999 to 2005. He has served as Chair of the RIBA Planning Group and a member of the RIBA Council. His blog The Gutter and the Stars can be read at http://pscpa.blogspot.co.uk/

Towards an Internet of things for the built environment

By Antonio Pisanò

The technology is ready. The question is: are policy makers capable of driving innovation or will it be down to the private market of system consultants to show the way?

At once fascinating and terrifying, the emergence of the Internet of Things seems an unstoppable drift that digital technology will impose on the way we experience the world.

For the ones who may not be familiar with the term, The Internet of Things is a digital image of the world.

In other words, every entity (product, system, person etc.) that forms part of the real world can have a digital counterpart in the Internet of Things. The key difference between the real world and the digital one is that the latter does not impose limits to the speed nor volume of interaction between entities. If humans are limited by their own specific capacity to receive, process and react to information, the entities forming the Internet of Things can exchange a large amount of data in a negligible amount of time. This makes the Internet of Things a responsive network in which systems interact and readjust themselves in relation to one another. The Internet of Things is a System of Systems. To give a simple - and rather simplistic - example, if the water distribution system could interact with the traffic management system, the latter could divert or reroute traffic to avoid the danger of driving through flooded areas as soon as the first system communicates to the second one of the presence of an issue. Humans can also divert traffic in flooded areas, but not as rapidly or efficiently. An in-telligent system of systems is faster and more efficient than human control. If we could introduce a systemic and responsive approach to the management of specific sectors such as food, energy, social welfare, housing or crime prevention, the use of resources would be improved, limiting waste and contributing towards a more resilient and balanced society.

Focusing on the impact on the built environment, how could we unlock the efficiency hidden in the integrated management of buildings and infrastructure together?

First we need to gather the data that will let us represent the built environment in the digital realm to create a digital image of the real system. The task is challenging to say the least. Some sectors, like the energy supply and property markets, already have robust and efficient data management systems, more or less ready to be shared; others, like the construction industry and the planning system are quite far behind. One solution could be to use planning applications to gather data in a standardised format across the country.

The online planning portal could quite easily become a digital building site.

Instead of planning applicants being asked to submit design and access statements, they should be asked to upload a single project database (similar to BIM databases, only expanded) containing relevant, standardised, information about: location, massing, building design, materials, carbon emissions, structure, services, energy performance, occupancy, programme, resources, jobs…etc..

If planners and applicants shared the same database structure, applying for planning would become as simple as uploading content to a wordpress website. This could also make the planning system faster and less over-loaded.

The key to achieving this is defining the specifications of such a database in order to capture the complexi-ty of the built environment: GIS data, to define the location of the building on the planet; geometrical data, to define its volumetric massing and deal with right of light issues; an overall 3D model of the borough, to allow swift evaluation of the impact of the proposed development on the Local Character, to ensure compliance with Conservation Area guidelines; and an articulated palette of facade materials, to simplify the evaluation of Local Amenity and Planning Conditions. Through this, Unitary Development Plans could not only be written or illustrated but built in a digital space to highlight opportunities for things like Public and Private Partnerships etc. Additionally, augmented reality simulations would allow planning committees to speed up their decisions avoiding the confusion often caused by drawings and renderings.

Our recommendation is to re shape the planning system as a digital building site through which applicants are asked to share their BIM database according to a standard form. This would make built environment cross-sys-tem integration more efficient.

The change we propose is at once revolutionary and conservative. If an expanded BIM model is a bit more refined than a design and access statement, filling in a form is still filling in a form!

A BIM based planning system will allow the creation of an Internet of Things for the built environment through which policy makers will be able to improve the efficiency of

•Housing:mappingoccupancyanddrivingcounciltaxvariation •Energy:mappingbillsanddrivingretrofitting •Employment:mappingunemploymentanddrivingdevelopment …etc.

The technology is ready. The question is: are policy makers capable of driving innovation or will it be down to the private market of system consultants to show the way?

Marcel Mauer Architecture work in partnership with CAIRE Urbanistica to find innovative, data driven solutions to improve the outcome of public and private planning, from regional strategies to building design. AuthorAntonio Pisanò is an architect, co-founder and director at Marcel Mauer Architecture (UK) and partner at CAIRE Urbanistica in Italy, leading the Smart City sector. Antonio has vast experience both in the UK and overseas, master-planning residential, hospitality, office and mixed-use projects in the UK, Turkey, Italy and China. Before setting-up Marcel Mauer Architecture he worked as designer and sustainability champion at Sheppard Robson Architects. His active contribution to Government Policies and the debate around innovation is the construction industry is proven by his membership of the CIC (former BIS IGT) 2050Group and the co-chairmanship of G4C – Constructing Excellence. He guest-lectured and took part in various events related to sustainability and smart city, published articles on architecture, planning, big data and contemporary art and took part in exhibitions, TV and radio shows.

Science, communication and urban planning practice

By Flora Roumpani and Prof. Sir Alan Wilson

The rise of the digital age provides a unique opportunity to achieve real interdisciplinarity by integrating our knowledge – ‘urban science’ – with an ability to communicate, to connect planners to urban communities.

When the architect Le Corbusier and his peers from the 20th century proclaimed that “form follows function” they expressed a significant idea which was widely applied in the architecture of the industrial revolution. In many ways this notion also applies on the urban environment. Cities are complex systems and their spatial substance is heavily dependent on socio-economic activities. This complexity is one of the reasons why urban planning practice has always been supported by a range of disciplines. The rise of the digital age provides a unique opportunity to achieve real interdisciplinarity by integrating our knowledge – ‘urban science’ – with an ability to communicate, to connect planners to urban communities. Figure 1 outlines the steps through which this could be achieved by harnessing big data and technology.

The first step is the digitisation of various forms of data, as outlined on the left hand side of the diagram. The data can vary between static data, dynamic real time data (i.e. ‘big data’) and planning knowledge. If processed appropriately, these datasets can be put into an intelligently searchable information system to provide inputs for a wide range of analytical and modelling software. Based on these inputs, different planning scenarios can be generated via advanced mathematical techniques developed to model the high levels of interdependence between the elements of an urban system (Wilson, 2013). ‘Spatial interaction’ and ‘predictive modelling’ are examples of these techniques.

Visualisation helps connect function (the science) with form (design).. The emergence of procedural modelling in computer graphics has made it possible to create 3D visualisations of possible scenarios based on urban theory rules (Parish, Müller, 2001). This enabled generating and managing digital representations of not only physical, but also functional characteristics of a facility. For the built environment, this means that fully realistic 3D planning scenarios of cities can now be automatically generated using urban modelling (Roumpani, 2013).

Figure 1. An interactive modelling and planning system

Such interactive visualizations embed communication at the heart of modelling systems, enabling products like ESRI’s City Engine to engage with urban communities via interactive applications and online platforms (e.g. webGL platforms).

Early implementations of these include projects that attempt to explain the interaction between different land uses of a city using spatial interaction and related techniques (Figure 2).

Figure 2. Polycentric radial town example, demonstrating the basic location theory by connecting geographic location with economic activity. Colours in the 3D diagram indicate different activities. Land uses with higher economic activities tend to concentrate closer to the 3 centres of the city shown in red.

These projects demonstrate how dynamic urban modelling theories can be used to model 3D real-time interactive animations of cities. They offer users the ability to test how cities may evolve under different scenarios, which they can control and alter. The outputs are both visual and analytical, as there is the option of providing 3D diagrams and matrices of different statistics (Figure 3).

Figure 3. Interactive controls of the model’s indicators from City Engine.

The development of such systems provides a framework for the digitization process, allows the integration of planning inputs and models, and communicates the outcomes to different groups of users.

A challenge for digitising the planning system will be defining spatially-related problems that accurately reflect real life, and use these as a basis for modelling in a planning context (Smith et al, 2014). If this can be achieved, a Digitised Planning System could provide a means for testing the impact of various developments and improve the ability of local authorities to interact and communicate with the public. For example, it could help determine optimal social infrastructure and accessibility provision within new developments, by allowing users to experiment with, visualise, and simulate the impact of different locations for retail, education or health facilities, and compare this with optimum results according to urban theory. See Figure 4.

Figure 4. Allocation of facilities using urban modelling and accessibility measures. The user in this case can interactively select the areas that generate employment (red) and the interactive system outputs the optimum locations for social infrastructure such as retail facilities (Shown in yellow).

In summary: the challenges for digitising the planning system are fourfold. Firstly, integrating new data sources with old ones as in Figure 1, and to have a means of visualising the data as in City Engine (Roumpani, 2013); secondly, making the best use of this data with analytical and modelling tools (cf. Wilson, 2013); thirdly, integrat-ing these methods with existing approaches to master planning (Kropf, 2013); and fourthly, to demonstrate how these methods can be communicated to the wider community and thus form the basis for interactive planning.

ReferencesKropf, K. (2013) Intelligent Master Planning Tools, Working Paper Series 194, CASA, UCL.

Parish Y. I. H., Müller P. (2001) “Procedural Modeling of Cities.” In Proceedings of the 28th Annual Conference on ComputerGraphicsandInteractiveTechniques,301–308.SIGGRAPH’01.NewYork,NY,USA:ACM.

Roumpani, F. (2013) Developing classical and contemporary models in ESRI’s City Engine, Working Paper Series 191, CASA, UCL.Smith A.H., Batty M., Hugel S., Roumpani F., Gray S. (2014, under review) Self-monitoring, Analysis and Reporting Technologies (SMART) in Cities: Data, Dashboards and Procedural Urban Modelling. The Cambridge Journal of Regions Economy and Society.

Wilson, A. G. (2013) the science of cities and regions, Springer, Heidelburg.

AuthorsFlora Roumpani is an MRes graduate and a PhD candidate at the Centre for Advanced Spatial Analysis in Bartlett UCL and holds a diploma on Architecture Engineering from the Department of Architecture in the University of Patras. During her studies she worked as a researcher in the Laboratory of Urban and Regional Planning in research projects relating to urban analysis and visualisation. For 4 years, she worked as an architect as part of the urban planning team in Doxiadis Associates, in several projects in Greece and abroad. Research interests include issues concerning the future of the city, virtual environments and urban modelling. Research blog: www.en-topia.blogspot.co.uk.

Sir Alan Wilson FBA, FRS is Professor of Urban and Regional Systems in the Centre for Advanced Spatial Analysis at University College London and until recently, was Chair of the AHRC. His current research, supported by ESRC and EPSRC grants of around £3m, is on the evolution of cities and the dynamics of global trade and migration. He was Vice-Chancellor of the University of Leeds from 1991 to 2004 when he became Director-General for Higher Education in the Department for Education and Skills. He is a Fellow of the British Academy and of the Royal Society and was knighted for services to higher education in 2001. His book, Knowledge Power, was published in 2010, The science of cities and regions, and his five volume (edited) Urban modelling in September 2012.

Envisaging a digitalisedplanning system

By Peter Insole

I see a digitised planning system as primarily an innovative web-based tool through which local communities can engage in shaping neighbourhoods and learn about the historic development of places; and use this knowledge to inform planning decisions at the neighbourhood scale.

The planning system should enable local authorities to promote the creation of quality places, encourage greater participation in placemaking, and embed a range of aspects - such as heritage- at the heart of sustainable urban design. Big data and smart technologies provide new opportunities to achieve these objectives.

I see a digitised planning system as primarily an innovative web-based tool through which local communities can engage in shaping neighbourhoods and learn about the historic development of places; and use this knowledge to inform planning decisions at the neighbourhood scale.

In addition to standard planning data, a digitised planning system should therefore provide access to local historic archives through a mapping interface allowing users to overlay different types of maps (spatial and temporal) and to upload locally sourced information. This would directly help enhance important Local Authority records, e.g. Historic Environmental Records (HER), and make them an immediate material planning consideration.

A digitised planning system should also provide an online facility to map community character to enable members of the public to define the character and distinctiveness of their neighbourhood. It should be capable of creating a visual language that links the character’s description to local development management policies on local character and distinctiveness.

A range of neighbourhood planning tools already exist that help assess the qualities of a place, showing what improvements are needed, and focusing people to working together to achieve them e.g. Placecheck. However, a digitised planning system should go beyond these applications in having a direct and systematic link between the participatory process and planning and policy.

At Bristol City Council, we have established Our Place – a tool that enables communities to participate in character mapping to understand the value of the process, and define their local context in accordance with HER data structures. This approach reduces the resource implications of a participatory approach for the local authority as the data collected is returned in the appropriate format and can be directly related to Development Management policies on Local Character and Distinctiveness.

We have trialled Our Place in five varied areas of the city - from inner city environments to open parkland, and conservation areas to post war housing estates. In one instance Our Place has enabled the community to create a Conservation Area Character Appraisal with limited resource implications for the local authority. After one day of mapping by the community, the results were added to an established Our Place character template that defines the local character areas along with specific challenges and opportunities. Other communities have

used the approach to inform neighbourhood planning, and to feed into the context study for a Bristol Central Areas Plan.

Applications like this could be scaled up to create a successful digitised planning system that saves costs and money, and provides a platform for linking community participation to planning policy and guidance about the value of local places and their character. It could help create community-led incentives to inform community design statements, like in the case of Our Place. But it could also be used for many other projects to drive the creation of sustainable and socially productive places. Additionally, it could facilitate the planning process by quickly disseminating information (e.g. share draft documents and invite comments) and record and evaluate the process through social media. This would help share the community’s experience and encourage others to participate. These type of initiatives will have to be reliant on specialist local authority data managers to help create an efficient and effective system.

The approach would link the public back with the planning process, and is likely to appeal to local ward councillors who may see a digitised planning system as a way for local amenity groups to become proactive in shaping the future of their neighbourhood in partnership with their local authority.

If the already existing neighbourhood planning engagement tools, such as Our Place, could be rolled out to an LA level – and beyond – the results of neighbourhood-level planning projects could be seen in relation to each other, which would help widen the understanding of the distinctiveness of individual neighbourhoods and begin a collaborative placemaking process based on a thorough understanding of place.

The tools and data to achieve a digitised planning system already exist – now it’s a matter of connecting all the dots and rolling out the approach across Local Authorities to bring back character and local meaning to place creation.

AuthorPeter Insole has worked for Bristol City Council since 2007 in their multi-disciplinary City Design Group consisting of urban designers, conservation officers, landscape architects and archaeological officers. During this time he has managed the Bristol Historic Environment Record (HER), provided archaeological development management advice and contributed to the creation of the city’s heritage planning policies. In 2010 Peter successfully applied for funding to develop the web resource Know Your Place (www.bristol.gov.uk/knowyourplace). This unique resource now underpins the council’s approach to the historic environment making archives more accessible and encouraging members of the public to share their own understanding of place.

A SMART approach to digitalplanning and design

By Tim Stonor

Fundamentally, the ways in which built environment data is used must change. Planners need to think at a much finer grain than before and architects at a broader one.

From pedestrian precincts to cul de sacs and upper level walkways, many innovations in urban planning and design have been launched with great optimism, only to blight new developments with massive social and financial costs. I believe there are two key reasons why this process of trial and error continues to happen: first, the scarcity of real knowledge about how people behave and, second, a shortage of accurate and reliable forecasting tools to test plans in advance. Until recently it has been expensive and time-consuming to overcome these issues: teams of observers with clip boards are costly; transcribing video is time-consuming. However, the rise of the “smart” era has witnessed an explosion of data capture and analysis techniques that can give us accurate and useful insights into how people behave. This matters because professional failure creates public concern.

If using big data effectively to design places can result in developments that work better for the public and local economy, then local authorities should not only seek to use the best analytics to capture it themselves, but should also demand the same of the private sector. But how can they and others use the newly available data effectively?

Fundamentally, the ways in which built environment data is used must change. Planners need to think at a much finer grain than before and architects at a broader one.

At present, architects use Building Information Modelling (BIM) systems that handle data at the building level. While BIM can stretch to small clusters of buildings, it does not usually allow buildings to be set in their wider urban contexts. As a result, the important influence of context on place is lost and too many buildings are designed in isolation, with obviously negative results once built.

Planners on the other hand tend to work from regional and city-wide scales down to ward and postcode levels, where their engagement with urbanism stops. But this can prove too crude to get an accurate picture of what is going on at the important human scale. What they need is to be able to analyse data to inform decisions - such as transport plans or changing land values – at least down to the level of the individual street segment and ideally to the different buildings that make up the street.

A digitised system of planning and design should allow all of the buildings to talk to each other, then all the blocks in a neighbourhood to talk with each other, then all the neighbourhoods within a district to talk to each other and so on. This would be an “Urban BIM”: a system that integrates professional activity and leaves no spatial voids.

But what does “talking to each other” actually mean? It certainly involves visualisation of data on a common platform. But it also means going beyond visualisation into data analysis, correlation and modelling. I am troubled by many of the conference presentations and discussions about smart cities that focus, sometimes obsess, on the visualisation of data - the creation of pretty maps and video clips - then go no further. A Digitised Planning System should be able to understand relationships between a number of different issues combined

together, as the most important potential benefits desired by planners or developers are likely to be the resultof a combination of these. In my own experience this means being able to associate “input” decisions on spatial layout and land use to “outcome” phenomena such as land value, movement, crime risk and carbon emissions.

A Digitised Planning System should equip architects, planners and stakeholders generally to properly weigh up the pros and cons of different options in delivering outcomes. I offer the following SMART approach:

A SMART Approach to Digital Planning and Design

Sense/Survey Capture useful “urban performance” data such as the demographics of a particular place, location of different types of retail, types of employment and typical travel patterns as well as “urban form” data including spatial accessibility, topography, building location, capacity and condition.

Map Spatially visualise that data e.g. develop maps that geo-locate the various urban performance and urban form characteristics.

AnalyseUse statistical tools to search the data for patterns, associations and correlations e.g. link observed pedestrian movement data with spatial accessibility levels and factor in the land use attraction created by shops and transport nodes. Infer via a software model simulation where residents are likely to want to travel to in the city and what sort of uptake there might be for a new bus route or cycle path. Use that software model to try out different options for changing the area and review how they would impact on the way the city works, in order to decide on which one would be most appropriate.

ReactProduce evidence-based policy, plans and detailed designs.

TestUse models to forecast the impacts of proposals in advance. Use the results of these forecasts to discuss ideas with stakeholders.

Once a particular option has been decided on and implemented, monitor how accurate the predictions were, in order to help refine and further develop the model. In other words, repeat the SMART cycle through further sensing, mapping, analysis, reaction and testing.

By taking such an approach, a Digitised Planning System would equip local authorities with their own live models of how their areas work across a range of scales, and use these to evaluate the likely impacts of designs from developers on the wider city. This then would make it possible for evidence-based planning decisions to be taken, giving local authorities firm grounds, for example, to negotiate design changes with developers.

One example of this approach is the City of London’s current development of a model to describe pedestrian movement within the city that will enable it to test how particular development proposals would affect this. Much experience has been gained over the last few years in the methods and tools required to develop models such as that sought by the City of London: models that are comprehensive enough to enable the overall impacts of different options to be thoroughly and reliably tested against each other yet sufficiently detailed to inform architectural and landscape design discussions.

My belief is that this approach can lead to the creation of a coherent Digital Planning System in the UK, which reconciles forecasting with accuracy, and the public with the planning process.

Tim Stonor is an architect and urban planner. He is an internationally respected expert in the analysis and design of human behaviour patterns in buildings and urban areas. His work explores how social, economic and environmental value is created by the movement, interaction and transaction of people in space. Tim advises public, private and community organisations worldwide. His approach combines robust analysis and visionary thinking. He is Managing Director of the strategic consulting firm Space Syntax Limited, which he founded in 1996. A director of The Academy of Urbanism, Fellow of the Royal Society of Arts, winner of the prestigious Harvard Loeb Fellowship and Advocate for the EPSRC, he is a Visiting Professor at University College London. He recently joined the Lead Expert Group of the UK Government’s Foresight project on The Future of Cities. Tim also speaks regularly at conferences throughout the world.

Author

Convergent City: Imaginingplanning in a digitised future

By Ulysses Sengupta and Robert Hyde

Society is changing due to new digital technologies and planning can help define this transformation.

Any discussion about digitising the planning system is meaningless without discussion on the future context of operation and aim. This report recognises new digital technologies as a significant future disruptor (Manyika et al., 2013) across disciplines. In order to address this change proactively there is a need to engage with potential futures.

Working with future trajectoriesWhile it is impossible to predict the specific outcomes of future technologies, current trajectories strongly indicate increased digitisation, increasing symbiosis between people and technology and the increased use of machine learning to perform tasks – such as weather pattern prediction or gene sequence analysis - requiring multiple iterations and calculations that would manually prove impossible. We may not know the future, but examples such as the governments’ Open Data initiative clearly demonstrate direction, and we can choose to approach this direction in a flexible and adaptable manner.

Machine learning will enable city simulationMachine Learning, a branch of Artificial Intelligence (AI) (McCarthy et al., n.d.), is the study of systems that learn from data. With the advent of Big Data, this branch of research has come to the forefront. Applications using machine learning already surround us. Google or Bing search engines rank websites for our web searches based on relevance, and junk mail filters learn from our habits in order to filter more efficiently. Pattern recognition - the ability to recognise or assign a value to a new input – is an essential part of machine learning, and recent developmentsincomputervision(E.g.facerecognition)andNaturallanguageprocessing(NLP)demonstratedirect translation of real world knowledge into digitally recognisable data without the need for prior categorisation. Machine learning can now create new systems that can regulate themselves based on external references, enabling the development of simulated cities that reflect real ones. The parallel behaviour of virtual/simulated cities and real ones will depend on AI research addressing the collective intelligence we see in cities (Weinstock and Gharleghi, 2013).

Cities are complex adaptive systemsCities are complex adaptive systems (CAS). I.e. systems that exist without a singular form of top-down control and evolve over time at multiple scales through an ability to learn. Other examples are the stock market, the biosphere and the ecosystem, the immune system and most human social group-based endeavours in a cultural and social system such as political parties or communities. Parallel computing and machine learning are providing new possibilities for digital simulation of these emergent systems. CAS uses internal models to ‘anticipate’ the future (Holland, 1992). I.e. test multiple future scenarios and adjust current actions based on learning. Digital simulations will provide decision making tools for people, planning and governance to test actions in the context of resilience (Walker et al., 2004)(Holling, 1996) and adaptation.

The convergent cityThere is an increasing convergence between the virtual and the real. This is manifested in machine learning, simulations of real world systems and the growth of the Internet of Things (Ashton, 2009). This convergence in the context of digital simulation models of cities is an essential future trajectory leading to increasingly accurate

simulations and automatic data exchange between the real and the simulated/data city. Sensors and monitoring devices installed in the name of Smart Cities may have a role to play in data acquisition, but the real potential is in the interfaces allowing interaction with these AI enabled data models.

Open data and direct democracyOpen data promises a new phase for society where people and organisations at multiple levels will access customised services in exchange for voluntary data surrender. Resilient and experimental socio-economic endeavours will result from increased flow and legibility of information allowing maximum awareness of context and changes. In the design and management of our environment, the potential of Open data lies not only in access to current information, but in direct forms of democratic city planning using interactive interfaces. Instead of having the option to object to top-down planning decisions only, citizens will be involved in a bi-directional process enabling suggestion of ideas and voting on projects and ideas they desire most. Structural change resulting in an accountable government providing feedback and real actions along with a digital simulation models to test new ideas will both be essential to such an outcome.

Digital planningSociety is changing due to new digital technologies and planning can help define this transformation. Structural shifts afforded by digital disruptors requires an interface between people, governance and the environment, more than increased efficiency of access to archival information. Digital planning must be the portal through which a) a virtual city is updated in real time; b) data is visualised in recognisable geo-spatial form; c) simulations are run to explore the viability of existing/future policies and interventions; d) urban debates and discussions take place; e) the real and virtual cities are convergent. The question of what this future interface looks like must be open-ended. However, given that we experience the complex interactions of real cities in four dimensions (including temporal), this is a useful cognitive starting point for a virtual city (with additional data displays) platform. A convergent virtual city that is updated in real time, incorporating machine learning and utilising simulations to test future scenarios to allow informed decisions and identify current and future problems and opportunities. Once in place the digital planning system must act as an open platform encouraging Civic Hacking for individuals to develop additional customised interfaces and services and become evolutionary itself.

AuthorsUlysses Sengupta is a Senior Lecturer at the Manchester School of Architecture and was previously at the University of NottinghamandtheUniversityofEastLondon.Heworkswitha complexity science framework to address complex urban situations produced by the rapid rate of urbanisation today and the resulting extreme changes to the physical fabric of many cities. His research is interdisciplinary and overlaps with Future Cities, Smart Cities, Big Data and Open Government agendas. Ulysses’ current research focuses on how to design and manage future cities through co-productive platforms based around real time geo-spatial systems. He is also Director of Softgrid Limited a research, design and consultation practice, specialising in city planning, urban regeneration, computational methodologies and integrated approaches to sustainability.

Rob Hyde is an Architect and Senior Lecturer at the Manchester School of Architecture where he runs a Post Graduate Studio + Research Atelier and is the strategic lead on Professional Studies. With a particular interest in Future Cities, trans-discipli-nary collaborative working and application of complexity science onto the urban realm. His current research focuses on the future alternative physical/ spatial, business/practice and governance opportunities afforded by Big/Open Data, Smart Cities etc. – asking the question: What does policy look like? [or could/ should look like?] and developing platforms to facilitate this.