Listen to the episode:

You can find the shownotes through this link.

---

Are you interested in how large-scale urban modelling going beyond the usual land use and transport interactions?

---

Our summary today works with the article titled A new framework fro very large-scale urban modelling from 2021 by Michael Batty and Richard Milton, published in the Urban Studies journal. This is a great preparation for our next interviewee, Michael Batty as he is an expert on urban data management. Since we are investigating the future of cities, I thought it would be interesting to see how the urban models can be expanded. This article presents a new framework including extensive spatial systems, population, employment, and distributed social interactions, which allows the running of hundreds of ‘what if?’ scenarios

[intro music]

Welcome to today’s What is The Future For Cities podcast and its Research episode; my name is Fanni, and today I will introduce a research paper by summarising it. The episode really is just a short summary of the original paper, and, in case it is interesting enough, I would encourage everyone to check out the whole paper. Stay tuned until because I will give you the 3 most important things and some questions which would be interesting to discuss.

[music]

As soon as computers came to existence, they were tasked to investigated large-scale problems, like predicting human futures. Big data and big computation have gone hand in hand, and although big data seems decreasing today, large-scale computation is in fact equally important. Computation generates more data that becomes even bigger, and the continued interrogation of data is what is big, not necessarily the data itself.

Urban models were first constructed in the late 1950s as simulations of land-use location and transportation flows. The limitations of the early decades of simulations have greatly decreased as miniaturisation has continued and it is now possible to operate quite big models on personal devices. Models can be run on smartphones so computation is no longer a severe problem. The model’s use of supporting strategic planning has also not changed. The ‘what if?’ investigations are still in the centre of large-scale urban modelling to inform predictions and explore scenarios for future cities.

The main benefit of current models is their fastness. Numerous solutions can be run in rapid succession to investigate specific decisions and scenarios. The article focuses on outlining a new model clearing the way how users can interact with such frameworks across a variety of dimensions. The proposed framework is essentially geared to choice models which are static simulations of urban structure at a single point in time. This static approach makes the model run faster and allow many datasets, thus viewpoints to be included.

Data in cities is spatially extensive and can never be detailed enough in terms of how many different types of sectors and population are characterised. Computational limits have been lifted with the advancements of technology. The quest to build larger urban models has thus revolved around disaggregation of activities, spatial units and temporal intervals to the finest categories possible. However, the emerging models are even worse for planning than their predecessors. Models are hard to really connect to reality in spite of real time data or cellular models.

The authors’ model focuses on urban flows at a cross section in time, which simplifies their approach. Their model is essentially a sketch planning tool to enable ‘what if?’ type impacts on the location of employment, population and transportation infrastructure to be evaluated. The model needs to be able to trace the impacts of infrastructure changes such as national and regional high-speed rail lines, new motorway systems, airports, national parks policies, green belts, large-scale housing developments and so on. The model also needs to be available to any user for any place within a region at any time – so it must be free, easily accessible and usable. These all brought the authors for an urban modelling as a web service solution. The model is called QUANT which stands for Quantitative Urban ANalytics forecasTing.

The single biggest problem in urban modelling is drawing the line between what is modelled inside the system and its wider environment. This problem is most severe when it comes to spatial definition. As the world has become global, it is increasingly difficult to identify such lines between what is important to a model within any particular city and what can remain as a passive input from the wider environment, the rest of the world. The authors used population data from census, road bus and rail network, and walking zones. Other data, like land area, greenbelts and floorspace, among others, are input, and these are available as options for measuring residential attraction.

As an example, they used the model on the UK, meaning England, Scotland and Wales. First they used it to visualise the spatial systems of these countries. This investigation revealed population density and employment connections, accessibility issues, and the connected transportation challenges. The framework exploits the power of modern computation in ways that enable us to handle compute-intensive models, and big data, which is big because of its spatial extent, the relationships between its elements and its continued use and reuse. As the modelling framework is available to anyone at any place at any time, it can generate very large numbers of users and scenarios investigated by those users.

To investigate the future of cities, some systematic and massive search is required. We can only generate hundreds of alternatives within evolutionary design process that continually improves the search for ever-better urban future if we use ever-faster models which enable us to proceed this way. There are now techniques for exploring such spaces. The models need to be embraced by users who have enough skill and awareness to be able to use them in planning support to generate plans to test ‘what if?’ style predictions. This remains one of the greatest challenges in using big data and large-scale models in a way that links good theory to good practice.

[music]

And a summary from ChatGPT: This article introduces a new framework for very large-scale urban modelling that is able to simulate the dynamics of urban systems in real time. The framework is based on a set of computational modules that can be combined to represent different aspects of urban life, such as transportation, land use, and social interactions. These modules are designed to be scalable and can be combined in different ways to model urban systems at different scales, from individual buildings to entire cities. The article discusses the potential applications of this framework in urban planning and policy-making, and highlights the need for new methods and tools to address the complex challenges facing cities in the 21st century. Overall, the article provides an important contribution to the field of urban modelling and highlights the potential of new technologies to improve our understanding of cities and inform more effective urban policies.

[music]

---

What was the most interesting part for you? What questions did arise for you? Do you have any follow up question? Let me know on Twitter at WTF4Cities or on the wtf4cities.com website where the transcripts and show notes are available! Additionally, I will highly appreciate if you consider subscribing to the podcast or on the website. I hope this was an interesting paper for you as well, and thanks for tuning in!

---

[music]

Finally, as the most important things, I would like to highlight 3 aspects:

1.  As technology progresses, our abilities to model urban scenarios widen to test and investigate different ‘what if?’ scenarios in urban decision-making.
2. The models enable us to create evolutionary design processes that continually improves the search for ever-better urban futures.
3. The models need to be embraced by users who have enough skill and awareness to be able to use them in planning support to generate plans to test ‘what if?’ style predictions.

Additionally, it would be great to talk about the following questions:

1. What do you think about urban modelling?
2. What do you think what data should be involved in such urban modelling tools?
3. What would you like to investigate with such models? What would be your ‘what if?’ scenario for your city?

[outro music]

---