AI will manage data-driven modern cities in real time, says MU researcher

Imagine a world where all the information about the city or town you live in – from bike share availability to flood detection to nocturnal bat activity – is readily available in a 3D map on your mobile phone. A world where data-driven cities use geospatial artificial intelligence, such as interactive mapping, 3D environments, modelling and simulations, to manage their urban environments and tackle problems in real time.

The digital transformation of maps and mapping that has been underway in recent decades is far from over and is likely to see highly visual 3D maps replace 2D maps among other advances, according to a presentation by Prof Rob Kitchin of Maynooth University’s Department of Geography and Social Sciences Institute (MUSSI) to the 2024 International Geographical Congress (IGC).

He was delivering the keynote address at a session on the future of mapping at the largest gathering of international geographers which is taking place in Ireland for the first time. In his talk, Prof Kitchin said the process of digital transformation continues to evolve and will demand new ways to make sense of mapping.

Prof Kitchin said that urban ‘digital twins’, a twinning of physical environments and systems and a virtual counterpart, are still in their infancy. Presently they exist as City Information Modelling (CIM), where a detailed virtual model mirrors the urban landscape with accurate topology and topography and realistic buildings and landscape rendering and in which administrative and real-time data can be visualised.

However, such models are far from being a true digital twin in terms of changes in the real world being updated in real-time in the virtual model, and changes in the model creating an alteration in geographic space.

“Yet CIMs and the promise of digital twins have captured the interest of many stakeholders. Governments see them as a useful tool for urban management and governance, formulating and assessing urban planning, and the monitoring of city services and policies. Architectural, construction, and planning businesses see them as a means to develop and sell urban designs and plans.

“Researchers aim to use them to produce urban simulations and models and to facilitate participatory projects relating to local places, and civil society organisations are interested in creating new kinds of counter-mapping initiatives that might open up new political conversations about urban development,” Prof Kitchin said.

He noted that CIMs allow:

• users to be immersed in, and walk or fly through, digital landscapes
• display many kinds of spatial and environmental data
• experiment with building and urban designs such as altering building dimensions
• run various forms of simulations and models such as traffic movement
• act as a collaborative environment for discussion, design practice and operations

By doing so, they have the potential to enrich public understanding of prospective urban design in public consultation, facilitate participatory planning and improve communication of sustainability science, and aid real-time management of city infrastructures and space.

He also noted that digital twins and deep maps prompt critical reflection on the nature of mapping. While such methods are cast as “value-free, objective” they are instead “an engine, not a mirror” in that they change rather than just reflect the world. In this sense, they are far from being neutral tools but reflect the ambitions and aims of their creators.

The IGC session was co-hosted with the RGS-IBG Annual International Conference taking place in London. Also presenting at the same event were Luke Barrington, AI Product Lead for Google Dublin, and Celine Rozenblat, Chair of the United Nations Geospatial societies on Global Geospatial Information Management who discussed the development of Geo-AI.

In his presentation, “Earth AI: Models of the planet as seen from satellites to Street View and everything in between”, Google’s Barrington spoke about using geospatial observations to detect floods and wildfires around the world and alert affected users, reconstructing 3D models of cities from aerial and StreetView imagery as well as Google’s Open Buildings project, a dataset of building footprints.

See more stories here.