AUTONEWS

Smartphones in cars offer new way to monitor city road conditions

The smartphone method doesn't just detect bumps, cars feed footage into an app that builds a live map of road conditions across an entire city.

The Monash team led by Dr. Yihai Fang, at the ARC Smart Pavements Australia Research Collaboration (SPARC) Hub, has already trialed crowdsourcing with 22 student drivers sharing data from their phones. They now hope to expand the project in partnership with road authorities. The research is published in the IEEE Internet of Things Journal.

"The more data we get from different vehicles, phones and road conditions, the stronger the system becomes," Dr. Fang said.

"This could one day help create a city-wide road health map powered by everyday drivers."

Dr. Fang said around 25 vehicles were fitted with smartphones and driven along Melbourne roads over two months. The team used different types of cars and phone positions to mimic real-world conditions.

"Our research shows that smartphones, supported by deep learning models, can reliably capture road roughness data across different vehicles and mounting positions," Dr. Fang said.

Road authorities currently use specialized survey trucks fitted with laser equipment to monitor road conditions. These provide accurate results but are costly and usually only deployed once or twice a year.

Climate change and extreme weather are placing additional stress on road networks, creating a need for more frequent checks. Affordable tools like smartphones could help provide extra data between surveys, supporting smarter use of Australia's $15.8 billion annual roads budget, about half of which is spent on maintenance and renewal.

"By using cars that are already on the road, we can make monitoring more frequent and responsive. This could help spot problems earlier, before they turn into costly repairs," Dr. Fang said.

The system, created by a team led by Dr Yihai Fang at the ARC Smart Pavements Australia Research Collaboration (SPARC) Hub, uses data collected from mobile phone sensors to build a live map of road health. Drivers’ smartphones feed footage and motion data into an app, which applies deep-learning models to detect roughness and bumps.

The approach offers a cheaper and quicker alternative to traditional road surveys, which currently rely on laser-equipped survey trucks. While these vehicles provide highly accurate results, they are expensive to run and are typically only deployed once or twice a year.

“With climate change and extreme weather putting additional strain on our road networks, there’s a clear need for more frequent monitoring,” Dr Fang said. “By using cars that are already on the road, we can make monitoring more responsive and spot problems earlier, before they turn into costly repairs.”

In an initial trial, around 25 vehicles were fitted with smartphones and driven across Melbourne roads for two months. Different types of cars and phone positions were used to replicate real-world conditions. The study showed that phone sensors, when supported by machine learning, could reliably capture road roughness data across varying environments.

The team has also tested crowdsourcing, with 22 student drivers sharing smartphone data to demonstrate how multiple contributors can strengthen the system. Dr Fang said expanding the project with road authorities could eventually deliver a city-wide road condition map.

“The more data we get from different vehicles, phones and road conditions, the stronger the system becomes,” Dr Fang said. “This could one day help create a city-wide road health map powered by everyday drivers.”

Australia spends about $15.8 billion annually on roads, with roughly half dedicated to maintenance and renewal. Monash researchers say affordable smartphone tools could complement official surveys, providing valuable data between inspections and helping governments direct spending more efficiently.

Provided by Monash University