Abstract
Information on road surface conditions is important and helpful for road users such as automobile drivers, particularly in snowy seasons. In practice, the surface conditions depend greatly on the weather, road users, location, and other relevant factors. This report is concerned with the reliable detection of the surface conditions using tire noise from road vehicles. The road/tire noise emitted from moving vehicles varies momentarily depending on the road surface properties. Then, it may be possible to passively and easily detect the state of the road surface: i.e., dry, wet, or snowy state. To detect tire noise, we used a commercially available microphone as an acoustic sensor, which enabled us to easily reduce the cost and size in realizing a practical system for detecting road surface conditions. We propose a couple of simple detection methods to classify the conditions into several categories of state and to improve the classification accuracy. From the experimental results obtained in snowy areas, it has been demonstrated that an accuracy of up to about 81% is attained in predicting the road surface states using only tire noise data.
