Estimate Road Roughness Using Smartphone Response Data - A Semi-Supervised Learning Approach

The smartphone-collected vehicle response data is being used to estimate the International Roughness Index (IRI). Among the existing smartphone-based methods, the datadriven approach, which involves training a machine learning model, is drawing more attention. However, surveying the IRI using conventional methods is expensive and there is limited labelled response data. In contrast, there exists a wealth of unlabelled response data from road users. This scenario presents opportunities for exploring semisupervised learning (SSL) algorithms, which have been insufficiently researched in this domain. This study addresses this gap by applying an SSL framework to refine an IRI estimation model. Our results show that the SSLtrained model achieves a lower RMSE than the fully supervised trained model.