Smartphone apps can measure road roughness—but not interchangeably
Two apps agree on bumpy roads, but systematic bias limits swapping one for the other.
The study collected data over 1700 km across Austria, Hungary, and Romania under real traffic conditions, using two smartphone apps for IRI estimates, in-vehicle GNSS-IMU measurements, and passenger comfort ratings. The two apps showed strong correlation, but systematic bias prevents interchangeability. An inverse relationship between IRI and passenger ratings confirmed perceptual sensitivity to roughness, and positive correlations between IRI and vertical acceleration validated the link between pavement irregularities and vehicle dynamics. The results highlight challenges of using consumer-grade sensing and perception-based evaluation for low-cost road roughness monitoring as an alternative to specialized survey equipment.
- 1,700 km of real-world data collected across Austria, Hungary, and Romania
- Two smartphone IRI apps show strong correlation but systematic bias prevents direct substitution
- Positive correlation between IRI and vertical acceleration confirms physical link; inverse with passenger comfort validates perceptual sensitivity
Why It Matters
Low-cost road roughness monitoring via smartphones could democratize pavement condition surveys for developing nations and municipal budget teams.