AI-enabled autonomous driving for urban sustainability: a Nanjing central area case study

• First integrates XGBoost, GIS analysis and multi-scenario simulation to explore the long-term impact of autonomous vehicles on the spatial structure of Nanjing's central urban area. • Reveals the outward redistribution characteristics of urban high-density development and the regularity of 61.1% density growth in peripheral subcenters. • Clarifies that the interaction between intelligent transportation and infrastructure triggers the transformation of urban space from monocentric to polycentric. • Predicts the reshaping effect of autonomous vehicles on urban density patterns and land use intensity. • Enriches the theory of interaction between urban space and transportation technology, providing innovative cases for ecological city indicators. With the rapid advancement of artificial intelligence technology, autonomous driving is profoundly reshaping urban transportation patterns and spatial structures, which presents both new opportunities and challenges for achieving urban sustainability goals. This research explores how AI-enabled autonomous driving can enhance urban sustainability by intensifying land use and optimizing urban density. Taking the central urban area of Nanjing as a case study, this research constructs a traffic accessibility–floor area ratio (FAR) correlation model. By integrating multi-scenario simulations and machine learning approaches, it quantitatively analyzes the impact mechanisms of autonomous driving technology on urban spatial structure. The findings reveal that: (1) Autonomous driving significantly enhances overall traffic accessibility, driving a shift in FAR distribution from a monocentric concentration toward peripheral expansion; (2) The Xinjiekou core area maintains an ultra-high FAR (3.8–4.8), demonstrating the functional agglomeration resilience of traditional central business districts (CBDs) under technological transformation; (3) In peripheral areas such as the southern part of Hexi New Town, FAR increases by up to 61.1%, forming linear high-density development zones along rail transit corridors; (4) Autonomous driving demonstrates a notable compensatory effect in peripheral areas, expanding the development potential of traffic-blind zones. By integrating carbon emission estimation with accessibility equity analysis, this study verifies the sustainable potential of autonomous driving in reducing per-unit carbon emissions and mitigating spatial inequality. The findings provide quantitative evidence for understanding AI-driven urban spatial restructuring and offer strategic insights for the coordinated development of intelligent transportation and land-use policies.