Overview

• Monash engineers used phone accelerometers, gyroscopes, GPS and video to detect rough patches and build live maps of road conditions.
• The team validated deep-learning models that adjust for phone mounting and vehicle suspension, improving accuracy across diverse cars and setups.
• About 22–25 drivers collected data over two months on Melbourne roads, demonstrating reliable estimation of pavement roughness in real traffic.
• The researchers say smartphone data can fill gaps between costly, seldom-run laser survey trucks, offering more frequent checks as roads deteriorate faster after extreme weather.
• Discussions are underway with VicRoads and the National Heavy Vehicle Regulator, and the team is seeking partners to pilot deployments on existing public and commercial fleets; the core research was published in the IEEE Internet of Things Journal.