Over a million developers have joined DZone.

Using AI to Predict the Growth of Cities

DZone's Guide to

Using AI to Predict the Growth of Cities

This article takes a look at a recent study on how AI can be used to better understand how cities grow and evolve.

· AI Zone ·
Free Resource

Bias comes in a variety of forms, all of them potentially damaging to the efficacy of your ML algorithm. Read how Alegion's Chief Data Scientist discusses the source of most headlines about AI failures here.

Cities are some of the clearest and well-used examples of a complex system, and whilst we are certainly better than we have been at managing their growth, they are, to a large extent, unmanageable.

A recent study by a team of Spanish researchers at the Universidade da Coruna highlights how AI can be used to better understand how cities grow and evolve, at least in a vertical sense. The researchers use an evolutionary algorithm that's trained on historical and economic data of an urban area to predict how the skyline could look in a few years time. The method was successfully deployed in the Minato Ward, in Tokyo.

Biological Growth

The team believes that cities grow in a similar way to self-organized biological systems. As such, they developed genetic algorithms that have been inspired by nature to predict how the urban landscape will change in a city.

"We operate within evolutionary computation, a branch of artificial intelligence and machine learning that uses the basic rules of genetics and Darwin's natural selection logic to make predictions," the authors say.

"In this type of computing, a multitude of possible solutions to a problem are randomly combined," they continue, "and a selection system is choosing the best results. This operation is repeated again and again until the algorithms get the most accurate results."

The team has built upon previous work on genetic algorithms to create one that seems to be able to accurately learn the growth patterns of urban areas from historical data, both from the construction history of a city and its economic performance.

Put It to the Test

They tested their work on one of the most active neighborhoods in the world. Tokyo's Minato Ward is home to a raft of multinational companies and has seen a tremendous amount of building work that has transformed its skyline.

The data was used to create a number of maps and 3D representations of the neighborhood so that they could then predict the number of buildings, and their likely locations within the district. They projected the growth between 2016-2019.

"The predictions of the algorithm have been very accurate with respect to the actual evolution of the Minato skyline in 2016 and 2017," the author explains. "Now, we are evaluating their accuracy for 2018 and 2019 and it seems, according to the observations, that they will be 80 percent correct."

The system appears able to predict both the number of skyscrapers in an area and their specific location. Suffice to say, the algorithm was not predicting too far into the future, and any development in such a near timeframe will have an awful lot of planning work around it already, so the algorithm would need to predict quite a bit further into the future to be truly useful, but hopefully that's something the researchers will continue to work on.

Your machine learning project needs enormous amounts of training data to get to a production-ready confidence level. Get a checklist approach to assembling the combination of technology, workforce and project management skills you’ll need to prepare your own training data.

genetic algorithm ,cities ,ai ,machine learning ,artificial intelligence

Published at DZone with permission of

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}