Road Pavement Health Monitoring System using Smartphone Sensing with a Two-stage Machine Learning Model

Published: 12 March 2024| Version 1 | DOI: 10.17632/4hmyws7cbc.1
Contributors:
,
, James Loney, Andrea Visentin,

Description

The Road Pavement Health Monitoring System Dataset is a comprehensive collection of GPS and motion sensor data gathered through smartphone sensing for the purpose of assessing and monitoring the health of road pavements. This dataset is associated with the research paper titled "Road Pavement Health Monitoring System using Smartphone Sensing with a Two-stage Machine Learning Model," .

Files

Steps to reproduce

The iOS application developed by the author was installed on two iOS smartphones, iPhone 11 and iPhone 13, to test the sensors’ stability across different smartphone models and avoid potential hardware-induced errors. Vibrations and GPS coordinates were collected at sampling rates of 100 Hz and 1 Hz, respectively. To validate the robustness of the data, the collection process occurred along two different routes, involving two distinct types of buses: the minibus Mercedes Sprinter Traveliner and the two-deck bus Volvo B9T. The smartphones were securely affixed facing forward under the front seat on the bus lower deck, positioned as close to the vehicle’s suspension system as possible for optimal data capture. The experiment routes comprise various road conditions, including flat stretches, slopes, sharp turns, well-maintained roads, road bumps, and anomalies such as potholes. The routes are located in Cork, Ireland, covering approximately 6 and 5 kilometers respectively. Given that the buses encountered multiple stops, starts, and unexpected driving maneuvers as they entered the city centre, the data collection occurred under entirely uncontrolled conditions. Importantly, the driver was not informed about the experiment to ensure the authenticity of the data collected. On the other hand, manual inspection, primarily carried out by recording images containing metadata, was conducted to create a labeled dataset for the proposed two-stage machine learning training and evaluation.

Institutions

University College Cork School of Engineering

Categories

Artificial Intelligence, Civil Engineering, Civil Geotechnical Engineering

Funding

EU Commission Recovery and Resilience Facility

22/NCF/FD/10932

Science Foundation Ireland

12/RC/2289-P2

Licence