Abstract

Urban areas account for 75% of global carbon emissions and will house 70% of the world’s population by 2050, making sustainable urban planning critical for achieving UN Sustainable Development Goals (SDGs) 11 and 13. This paper presents NetZero, an open-source prototype employing Genetic Algorithms (GA) to optimize synthetic city layouts for near-zero carbon emissions while balancing population density, construction costs, and resident happiness. The system represents cities as grids with eight building types and employs a normalized multi-objective fitness function incorporating carbon emissions, livability metrics, economic constraints, and eight spatial planning constraints. Evaluated across grid sizes from 30×30 to 70×70 with populations of 30–100 candidate solutions over 500–2000 generations, and averaged over 30 independent trials (10 per grid configuration), NetZero’s GA reliably optimizes a synthetic carbon proxy, achieving a carbon ratio below 5% in 90% of trials (27/30). The Python-based prototype with Streamlit interface serves as a proof-of-concept tool for AI-driven spatial optimization while highlighting critical data gaps in standardized emission factors needed for real-world deployment.

Authors

Vikas Jha, Deepti Goyal
Satyug Darshan Institute of Engineering and Technology, India

Keywords

Sustainable Development Goals, Urban Planning, Genetic Algorithms, Carbon Emissions, Multi-Objective Optimization, Smart Cities