AI Can Design Cities

Kaifu Ren
Civic Analytics 2019
2 min readSep 15, 2019
AI can help us develop urban spaces

Cities are complex systems. City planning or city design is traditionally an arduous, iterative and arbitrary process among different players, and city plans made for the future are usually based on the study of archaic datasets which makes them inviable and infeasible during implementation. Planning and prebuild process for developers, architects, planners, engineers, and city users, in some cases, can take years, which makes the design and build industry among the lowest-performing sectors in terms of productivity.

Spacemaker, founded in 2016, is building the world’s first commercial platform that uses AI to design cities and helps stakeholders to make better and faster decisions. Planners and architects can input some parameters such as zoning regulations, traffic environment, and building codes for a particular site, and the platform can spit out hundreds of design options within just a couple of hours. Planners can optimise the output based on a number of factors, such sun exposure, wind ventilation, noise mitigation, green coverage, waste management, etc. Developers can review each layout to measure return on investment (ROI) and investment risk.

This platform can be scaled to understand and design cities in any part of the world once there are developed and established datasets on city parameters. As data scientist and analyst, our job is to collect and prepare massive amounts of data and build machine learning algorithms to train computers to model the ideal city where our people can live in a better life.

--

--