Analysis of bridge monitoring data using deep learning

Abstract

Monitoring of vehcles at bridges are studied at National Institute of Informatics (NII), using deep learning. Passing vehcles and lanes are detected with velocity, number of axles and weight estimation based on measurement of strain gauge, accelerometer and surveillance cameras. Accurate analysis results are obtained by preparing large training data sets. Data management system to accumulate and search information of measurement is also developed. Currently, data exchange platform to integrate diverse information is under development.