Role of big data in building sustainable and efficient cities

Artificial intelligence (AI) and big data are changing the way cities are planned, managed, and understood, with new research highlighting a decisive shift toward data-driven urban land use strategies. A new editorial study by Thomas W. Sanchez and Soheil Sabri reveals how big data and machine learning are reshaping the foundations of urban planning.

Published in Land, the study titled “Big Data in Urban Land Use Planning” brings together insights from seven research contributions across diverse geographic regions, including Sri Lanka, China, the United States, and Sweden. The work presents an overview of how artificial intelligence, remote sensing, and multi-source data integration are enabling more precise, dynamic, and scalable approaches to land use planning, while also identifying critical gaps in implementation, equity, and predictive modeling.

Multi-source data integration reshapes how cities are mapped and understood

The study identifies a growing reliance on multi-source data integration to capture the complexity of urban environments. Traditional land use planning has often depended on static datasets such as census records and administrative boundaries, which can quickly become outdated or fail to reflect real-world conditions. In contrast, modern approaches combine diverse data streams including satellite imagery, nighttime light data, points of interest, transportation networks, vegetation indices, and population distribution metrics.

One of the contributions demonstrates how combining these data sources can redefine urbanization patterns, particularly in regions where official statistics are limited or inaccurate. In the case of Sri Lanka, a big data fusion model revealed significant discrepancies between administrative definitions of urban areas and actual functional urbanization, uncovering both overestimated and underrepresented regions.

This shift toward dynamic data integration allows planners to move beyond static representations of cities and develop more responsive and adaptive models. By incorporating real-time or near-real-time data, urban systems can be analyzed with greater precision, enabling more informed decision-making across sectors such as transportation, housing, and environmental management.

No single dataset can adequately capture the multidimensional nature of urban land use. Instead, the integration of heterogeneous data sources provides a more comprehensive understanding of how cities function, evolve, and respond to policy interventions.

AI and machine learning drive automation and analytical precision in planning

AI plays a key role in advancing these data-driven approaches, with machine learning models increasingly used to automate complex analytical tasks that were previously labor-intensive or prone to error. The study highlights several applications where AI enhances both the accuracy and scalability of urban analysis.

One major development is the use of supervised machine learning models for land use classification. By integrating data such as areas of interest, population density, and nighttime light intensity, researchers have achieved highly detailed classification systems that distinguish between multiple land use categories with significant accuracy. Among the models tested, advanced algorithms such as XGBoost demonstrated strong performance, underscoring the growing effectiveness of AI in spatial analysis.

In another case, deep learning techniques are applied to analyze street-level imagery and evaluate urban environments based on factors such as accessibility, safety, and comfort. These methods allow planners to assess urban spaces from a human-centered perspective while maintaining the scalability of automated systems.

AI is also being used to explore the relationship between urban form and function. By analyzing street networks and points of interest, researchers have identified patterns in how infrastructure influences the distribution of economic and social activities. These findings suggest that urban morphology and land use are closely interconnected, with implications for transportation planning and economic development.

Additionally, automated tools are being developed to streamline environmental simulations. One study presented in the editorial introduces a method for generating microclimate models directly from remote sensing data, significantly reducing the time and effort required for urban environmental analysis. This innovation highlights how AI can bridge the gap between data collection and practical application, enabling planners to evaluate the environmental impacts of land use decisions more efficiently.

The study also acknowledges ongoing challenges related to model transparency, data quality, and potential biases embedded in training datasets. As AI becomes more integrated into planning processes, ensuring the reliability and interpretability of these systems will be critical.

Persistent gaps in equity, forecasting, and real-world implementation

While the integration of big data and AI offers significant benefits, the study identifies several critical gaps that must be addressed to fully realize their potential in urban planning.

One of the most notable gaps is the limited focus on predictive modeling. Although current approaches excel at analyzing existing conditions, there is less emphasis on forecasting future land use changes or simulating alternative planning scenarios. Given the rapid pace of urban transformation, the ability to anticipate future developments is essential for effective long-term planning.

The study also highlights concerns related to social equity and environmental justice. While data-driven approaches can enhance efficiency and precision, they may also reinforce existing inequalities if underlying data sources are biased or incomplete. For example, open data may disproportionately represent certain neighborhoods or populations, leading to skewed analyses and policy outcomes.

Another challenge lies in the translation of analytical insights into practical decision-making. Many of the studies reviewed focus on methodological advancements, with less attention given to how these tools can be integrated into governance frameworks, planning institutions, and policy processes. Bridging this gap will require collaboration between researchers, policymakers, and practitioners to ensure that technological innovations are aligned with real-world needs.

The study further notes that while open data and open-source tools are becoming more prevalent, barriers to adoption remain, particularly in resource-constrained settings. Ensuring accessibility and scalability of these technologies will be essential for their widespread implementation, especially in the Global South.