AI Framework Developed for Climate-Resilient Road Construction
Researchers at IIT Gandhinagar have created an AI-powered framework to design more durable, climate-resilient rigid pavements tailored to specific regional conditions.
Annual monsoon rains necessitate frequent road repairs, but thermal stresses also significantly contribute to pavement degradation long before the first drop of rain falls. Rigid pavements, constructed with concrete slabs, distribute heavy loads effectively and are increasingly used on highways, airports, and city roads. However, daily and seasonal temperature fluctuations create internal stresses within these structures, similar to how repeated cooling and warming can make chocolate grainy.
These thermal stresses lead to progressive fatigue damage in concrete pavements, ultimately shortening their service life. To address this, scientists at the Indian Institute of Technology Gandhinagar (IITGN) have developed a machine learning framework. This AI-driven approach aims to support the creation of region-specific, climate-resilient rigid pavements, potentially leading to more durable roads and reduced maintenance expenses.
The research team focused on Gujarat, a state with diverse climatic conditions ranging from humid coastal areas to hot inland regions. Current standards for designing rigid pavements, particularly in countries like India and Nepal, often rely on outdated temperature data from as far back as 1974. Furthermore, existing broad climatic zones encompass vast areas with varied weather patterns, failing to capture localized thermal behavior and compromising the accuracy of pavement performance predictions.
Recognizing that different parts of a state experience varying thermal stresses, the IITGN researchers collected hourly weather data from 126 land-based grid points across Gujarat. This data, sourced from the Copernicus Climate Change Service's ERA5 database for two distinct climatic periods (1961–1991 and 1992–2022), was used for thermal modeling of rigid pavements. The simulations incorporated different slab thicknesses (200, 250, and 300 mm) and surface albedo values (0.30 for conventional and 0.50 for cooler pavements) based on Indian Roads Congress guidelines.
This development is significant as it leverages AI and machine learning to move beyond generalized design standards for rigid pavements. By analyzing localized thermal data and environmental factors, the framework enables the creation of more resilient infrastructure. This approach is crucial for extending the lifespan of roads in the face of climate change and reducing costly maintenance, a vital step for transportation networks globally.
Edited by the news editor with AI from the original report — please refer to the original source.