Smart modelling reshapes city planning; it can foresee multiple scenarios, provide customised solutions and improve decision-making through the application of data. This section introduces the use of smart modelling in the urban context. It focuses on three aspects: 1) possible data sources, 2) tools for data processing and 3) applications of mapping and modelling. It also covers other aspects, including third-party participation and technical expertise for smart modelling.

A good amount of data is essential for smart modelling. It is worth noting that building a smart model does not always entail the collection of new data. In fact, historical maps have proved very useful in the creation of flood-prevention models. In Da Nang, Vietnam, the city government, as a first step, collected historical flood maps, as well as past data on sea-level rise and climate change. It then developed a set of hydraulic parameters to build a predictive model.[i] Using the model, the city simulated changes in flooding patterns in certain areas and amended its urban development plan to mitigate flooding risks.

The adoption of third-party data also advances smart-modelling development. For instance, institutions including the US National Aeronautics and Space Administration (NASA) and the European Space Agency (ESA) provide satellite imagery with resolutions sufficient for land-cover mapping.[ii] Combined commercial and public data sources help solve complicated land-use issues[iii] (for example, by identifying hotspots of severe soil erosion and comparing historical and current land pictures).

Once the required data are in hand, the next step is data processing. In addition to big data analysis, machine learning[iv] and artificial intelligence (AI) are increasingly being used to analyse the urban context in real time. These technologies simplify the processing of datasets made up of cross-correlated information, and can thus observe urban development patterns dynamically.[v] First, city planners teach computers what to search for in different sources of data, then use machine-learning algorithms to integrate the results with administrative data to identify, for instance, flood-prone areas or illicit land use.

In Ho Chi Ming City, Vietnam, the government adopted rudimentary AI tools to mine satellite data sources, then used machine-learning algorithms that taught computers to automatically detect and classify different land cover types and land use. The results were overlaid with the city’s zoning data to create a map tracking potential zoning code violations.[vi]

Another commonly used data-processing tool is Geographic Information System (GIS) technology. Using GIS, planners can analyse and visualise real-world geographical data on digital maps and integrate them with different geographic components based on their needs.[vii] GIS technologies, along with 3D simulation, are being used for building-height control, climate-change adaption and development planning.[viii] This is usually known as “digital twinning” – creating a virtual replica of the actual environment, including physical objects, behaviours and processes. This technology helps cities, or even individual utilities, enhance data capture and real-time visualisation.

The Swedish city of Uppsala, for example, wanted to preserve its natural habitats while addressing a housing shortage. [ix] To increase the sustainability of its development plan, the city government adopted an interactive 3D model built on a digital elevation map captured by Lidar scanning.[x] The model helped stakeholders visualise how traditional infrastructure could be built in harmony with nature.

Smart models provide cities with an intuitive tool for evaluating, planning and improving cities’ development plans. Not only can planners simulate urban scenarios and test alternatives in a virtual setup, they can choose to communicate urban planning information with stakeholders through the visualisation of data. By doing so, planners and researchers have more opportunities to cooperate and add value to cities’ smart model development.

In Boston, several agencies worked together to transform the city’s administrative data into a set of parameters, the foundation of its 3D smart model. Using the model, cross-departmental teams could establish shared projects and set common indicators. The 3D smart model was open to city planners and citizens, allowing private investors to locate suitable areas easily and to invest in real estate or commercial infrastructure.^[xii] In addition, based on the 3D smart model, the Boston planning and development agency and the Massachusetts Institute of Technology (MIT) successfully created an urban energy model that would facilitate energy savings and the roll-out of microgrids.^[xiii]