Roadlink
Smartphone crowdsourced road roughness estimation method. Developed a deep learning model estimates IRI using smartphone in realtime by onboard processing without calibration. Any driver can use their smartphone with app to contribute the society to estimate road network health status!
Roadlink is an iOS app that measures IRI in realtime using any smartphones. The key feature of this app is the calibration-free nature which can be used with any type of vehicle(i.e., sedan, SUV, minivan, and pickup truck), driving speed. It uses state-of-art deep learning technology to estimate IRI independent from vehicle mechanical characteristics. Currently, it is under preparation for release.
This work is US patent filed:
Jo. H., Jeong. J.H, and Ditzler, G. CONVOLUTIONAL NEURAL NETWORKS FOR PAVEMENT ROUGHNESS ASSESSMENT USING CALIBRATION-FREE VEHICLE DYNAMICS, US patent pending
Related publication:
Jeong, J. H., Jo, H., & Ditzler, G. (2020). Convolutional neural networks for pavement roughness assessment using calibration‐free vehicle dynamics. Computer‐Aided Civil and Infrastructure Engineering, 35(11), 1209-1229. (Top journal within Civil Engineering, 1/136, JCR)
Jeong, J. H., and Jo, H. Toward Real-world Implementation of Deep Learning for Smartphone-crowdsourced Pavement Condition Assessment (in review)
Here is the sample IRI data in the Tucson area measured by Roadlink