Sensor-based Pavement layer change detection using Long-Short Term Memory (LSTM)

Abstract During construction, pavement projects often suffer from a lack of progress certainty, which leads to cost and time overruns. The construction should be monitored in timely accurate manner provide prompt feedback ensure project success. However, current monitoring practices (e.g., data collection, processing analysis) are manual, time-consuming, tedious, inconsistent, subjective error-prone. previous research study was limited only incremental road measurement. This preliminary proposes novel sensor-based method identify layer changes during using series algorithm for the approach development automated as-built measurement construction. In this study, were collected generating various scenarios controlled environment by simulating ground vehicle equipped with laser ToF (time-of-flight) distance-ranging sensor. Subsequently, Long Short Term Memory (LSTM) utilized on feature detection as ‘layer up’, down’ not changed’ classify change. experimental result demonstrates 84.91% promising overall average accuracy change classification control data, confirming potential implementation suitability detect layers real projects. low-performance measures (low precision, recall F1 score) up down suggest further improvement enhance robustness proposed model. can extended automate validating case.