'Self-repairing asphalt' that will solve road collapses will be developed using Google's AI

A research team from Swansea University and King's College London in the UK has reported an innovative research result that developed a self-repairing asphalt using biomass and Google Cloud AI. The research aims to solve the problem of road potholes, which costs the UK £143.5 million (about 27.7 billion yen) per year.

How AI could make “self-healing” roads a reality
https://blog.google/around-the-globe/google-europe/united-kingdom/how-ai-could-make-self-healing-roads-a-reality/

AI-powered self-healing asphalt: A step toward sustainable net-zero roads - Swansea University
https://www.swansea.ac.uk/press-office/news-events/news/2025/02/ai-powered-self-healing-asphalt-a-step-toward-sustainable-net-zero-roads.php

This research is attracting attention as an innovative effort to improve the sustainability and durability of road infrastructure. The research team brought together experts in materials science, chemistry, and computer science to successfully develop a new asphalt technology that mimics the self-repair mechanisms found in nature.

Typically, the sticky black substance contained in asphalt, called bitumen , oxidizes and hardens over time, resulting in cracks.

To address this issue, the team incorporated hair-sized capsules into the asphalt, made from a porous, plant-based material that was filled with recycled oil-based reclaimed material.

The self-repair process begins with the appearance of cracks inside the asphalt. The physical force of the cracks triggers the capsules to burst, releasing the recycled material inside. The recycled material then reacts chemically with the hardened bitumen, restoring its flexibility. This chemical reaction softens the bitumen, allowing the cracks to be 'stitched together' and repaired.

The researchers say the entire process can be completed within an hour, restoring the structural integrity of the asphalt, and is automated, without human intervention, with a low environmental impact, as the materials used are bio-based and recycled.

In addition, the team said that in collaboration with Google Cloud, they developed a new data-driven model that allows for complex molecular-level analysis. The AI model accelerated atomistic simulations to better understand the oxidation process and crack formation mechanisms of bitumen, enabling faster and more cost-effective analysis compared to traditional computational models.

According to the research team, approximately 20 million tons of asphalt are produced annually in the UK, and the production process is a significant source of carbon emissions. The research team believes that this self-repairing asphalt technology will not only reduce road maintenance costs, but also play an important role in achieving the UK government's goal of zero carbon emissions by 2050, and emphasizes that collaboration between industry, academia, and government is essential to put it into practical use.