Abstract
Proper maintenance of the road network is crucial to preserving and improving citizens¿ quality of life.
Otherwise, the costs associated with the transportation of goods and people would increase, since poor
road conditions are linked to a higher risk of accidents, greater fuel consumption, and consequently, an
increase in greenhouse gas emissions.Current methods for assessing pavement conditions on roads involve
carrying out inspections with specialized equipment to evaluate both the state of the surface and driving
quality or comfort. This means that the administration cannot inspect the entire road network annually
due to technical and economic limitations. In this context, crowdsourcing data from connected and
autonomous vehicles (CAVs) with enhanced attributes emerges as an alternative to the use of
instrumented equipment. Unlike traditional inspection methods¿whose results largely depend on the specific track followed by the specialized equipment at the time of measurement and whose data
collection frequency is very limited¿the use of data from CAVs provides more reliable, near real-time
information on road conditions, since each point of the roadway can be monitored by hundreds or
thousands of vehicles.Therefore, the main objective of this work is to evaluate the feasibility of using
CAV data to assess pavement conditions through the International Roughness Index (IRI). To this end,
CAV-derived data have been compared with traditional inspection data, establishing different thresholds
to quantitatively and qualitatively define pavement condition based on vehicle data.
