Creating a resource cadaster for asphalt pavements
Background
Built environment stocks – building and transport infrastructure – provide essential services to fulfill basic human needs, workplace, mobility, and communication. Road infrastructure constitutes an integral part of built environment stocks. It connects and shapes human settlements, ensuring daily mobility of people, access to jobs, and distribution of goods. Materials accumulated in road infrastructure over their long-life span could potentially serve as future resource providers, which have been termed as anthropogenic material stocks. However, there lacks comprehensive and spatially-refined understanding to manage resources in a sustainable and cost-effective way. This project aims to develop an exchangeable knowledge and data platform to inventory ‘material passports’ of asphalt pavements, which will enable resource recycling and circular design in the pavement sector moving forward.
Research Questions
Hot to estimate the quantity and quality of resources stored in asphalt pavements using GIS and Machine Learning?
- SQ1: how to harmonize GIS data of pavements from different sources?
- SQ2: how to integrate climate, soil condition, and surface information from different sources in a spatially consistent way?
- SQ3: how to predict the material type and layer thickness using design-guided machine learning models?
Methodology
This project will employ GIS technologies to harmonize and merge spatial information of pavements from authoritative and crowdsourced databases to generate accurate and comprehensive GIS datasets for pavements in Belgium. Besides, inspired by established pavement design principles and engineering wisdom, we will harness supervised-learning algorithms and the Long-Term Pavement Performance (LTPP) database to train models that correlate pavement properties with key design-relevant factors (e.g., climate, soil, and traffic). The trained models will be used to predict the material type and thickness of asphalt pavements in Belgium. Our final goal is to build a resource cadaster for pavements in Belgium with high resolution.
This project is one thrust of the CircularPave project.