The article explores an innovative approach to assessing the condition of the road surface by analyzing the vibrations that occur during the movement of cars. The authors use data from mobile device accelerometers to detect different types of roadway damage. One of the key aspects of the study is the application of machine learning techniques to analyze large amounts of vibration data. The authors compare the results and performance of different clustering algorithms, such as K-Means, Affinity Propagation, Spectral Clustering, and Agglomerative Clustering. The results of the study demonstrate that spectral clustering is the most effective method for detecting road damage based on vibration data. The advantage of the proposed method is its relative simplicity and low cost, since ordinary mobile devices can be used to collect data. In addition, this method allows you to automate the process of road inspection and detect damage at an early stage, which helps to reduce repair costs. The authors emphasize that further research can be aimed at expanding the functionality of the system, for example, at identifying groups of damage and predicting their development. Another promising area is the integration of this system with other traffic management systems to ensure more efficient road maintenance.