AI in Smart Cities Infrastructure

Abstract

Smart cities leverage networked sensors, data analytics, and AI-driven automation to improve the efficiency, sustainability, and quality of urban infrastructure and services. As European cities accelerate digitisation under frameworks including the EU Mission for 100 Climate-Neutral and Smart Cities and the European Green Deal, AI applications across urban mobility, energy, water, waste, safety, and public services are transitioning from pilot projects to operational deployments at city scale. This study evaluates AI applications across six smart city infrastructure domains in twelve European cities (population 50,000-2,000,000): adaptive traffic management, smart grid and energy distribution, water network monitoring, waste collection optimisation, public safety and surveillance, and citizen services automation. For each domain, two to three AI systems are benchmarked against conventional management baselines using operational data from 2022-2024. Key findings: AI adaptive traffic management reduces average vehicle delay by 18.4% across 8 cities deploying it. Smart grid AI reduces grid balancing costs by 14.2%. AI water network monitoring detects leaks 12.4 days earlier than manual inspection cycles. AI waste routing reduces collection vehicle km by 16.8%. Computer vision public safety systems achieve 84.2% incident detection recall but raise significant privacy concerns documented through stakeholder analysis. LLM-powered citizen service chatbots resolve 68.4% of queries without human escalation. A smart city AI governance framework addressing privacy, equity, and democratic accountability is proposed.