Multi-scale GeoXAI analysis of electric vehicle charging points spatial equity in Singapore

Publication Type

Journal Article

Publication Date

10-2026

Abstract

This paper proposes a geospatial interpretability approach to study socio-spatial inequality in Electric Vehicle Charging Points (EVCP) accessibility. This framework combines machine learning and econometrics to examine the spatial heterogeneity and social equity of EVCP accessibility in Singapore. The findings reveal multiple mechanisms underlying urban infrastructure equity: 1) Urban morphology exhibits spatial heterogeneity that mature central urban areas demonstrate high accessibility through their organic and dense street networks; 2) Socioeconomic status is proven to be the fundamental driver of vertical inequality, making market-driven private residential communities better able to overcome physical limitations than public housing; 3) A mismatch between spatial coverage and service capacity exists, new town planning commonly suffers from a nominal equity trap of high spatial coverage coexisting with low service capacity. Future infrastructure investment needs to move beyond a spatial supply-oriented approach to ensure a balance between efficiency and equity during the energy transition.

Keywords

Electric vehicle charging infrastructure, Explainable AI, High density cities, Spatial heterogeneity

Discipline

Geography | Urban Studies and Planning

Research Areas

Integrative Research Areas

Publication

Sustainable Cities and Society

Volume

149

First Page

1

Last Page

19

ISSN

2210-6707

Identifier

10.1016/j.scs.2026.107833

Publisher

Elsevier

Additional URL

https://doi.org/10.1016/j.scs.2026.107833

This document is currently not available here.

Share

COinS