Optimizing urban grid layouts using proximity metrics

Research output: Chapter in Book/Report/Conference proceedingChapter

2 Scopus citations

Abstract

Artificial intelligence is increasingly being adopted to optimize, simplify, and extend operations in various areas of knowledge. Over the years, AI has been defined in many distinct but connected ways, encompassing different subfields and methods. Although in recent years, AI has been equated with machine learning, this chapter explores another AI paradigm, search, and optimization. In the urban design context, artificial intelligence has great potential to enhance decision-making and various assessment processes by performing complex iterations and carrying out predictions speedily and accurately. Proximity, in turn, plays a fundamental role in promoting urban dynamics, influencing city arrangements, and impacting the overall quality of life. Important urban features related to sustainable urban design practices such as transit accessibility, density, land-use diversity, and walkability, which are crucial for easing traffic congestion and improving public health, can be either assessed or explained by proximity indicators. This chapter investigates the utility of computational optimization techniques at the urban design scale, aiming to improve the performance of urban grid layouts according to the physical and topological proximity metrics. To this end, we employ evolutionary multiobjective optimization, a subset of evolutionary computation, and a generic population-based metaheuristic optimization algorithm in the generation of urban fabrics and analyze the outcomes of orthogonal and nonorthogonal grid typologies, including regular orthogonal grid, irregular orthogonal grid, irregular nonorthogonal grid, and Voronoi-shaped grid. During the generation of optimal urban grid layouts, computational tools measure the shortest physical distance between locations in a neighborhood and verify Space Syntax measures such as integration and connectivity. Our results indicate the advantages and drawbacks of each grid typology, identifying that orthogonal grids are more appropriate for car-oriented cities, while nonorthogonal ones are more suitable for walkable areas.

Original languageEnglish (US)
Title of host publicationArtificial Intelligence in Urban Planning and Design
Subtitle of host publicationTechnologies, Implementation, and Impacts
PublisherElsevier
Pages181-200
Number of pages20
ISBN (Electronic)9780128239414
ISBN (Print)9780128239421
DOIs
StatePublished - Jan 1 2022

All Science Journal Classification (ASJC) codes

  • General Social Sciences

Fingerprint

Dive into the research topics of 'Optimizing urban grid layouts using proximity metrics'. Together they form a unique fingerprint.

Cite this