Estimation of the pavement condition index via tyre cavity noise measurements and AI signal processing

Regular monitoring of the pavement condition is an essential feature to ensure both traffic safety and passenger comfort. In fact, pavement distresses directly influence traffic safety and passenger comfort and have severe negative effects on air and noise pollution of the vehicles. In this field local agencies commonly use the Pavement Condition Index (PCI), which is based on pavement distress types and their spatial extension, as a reference indicator of the pavement condition. Unfortunately, traditional methods to evaluate PCI can require an on-site ”walk and look” approach that are time-consuming, expensive and may request the closure of the road to traffic. To overcome such difficulties, in the last years different measurement methods have been developed and tested. In this paper an innovative methodology to estimate the PCI in urban context, permitted a dynamic and continuous acoustic measurement of tyre cavity noise (TCN) by means of a microphone inside the tyre. This innovative way involved the definition of new pavements classifying parameters. The new indicators seem to show a good agreement with the PCI values evaluated with the previous methodology. A possible improvement of the method will be tested by applying an AI signal processing to infer the PCI value from the TCN.