# Construction and demolition waste material library based on vision systems data

## TL;DR

This paper introduces a material library using vision systems to help manage construction and demolition waste more sustainably.

## Contribution

A new material library is introduced, combining infrared thermography and hyperspectral imaging data for CDW analysis.

## Key findings

- The library includes data from infrared thermography and hyperspectral imaging for material characterization.
- The data is standardized and compatible with machine learning tools for predictive modeling.
- The library supports AI algorithms to improve CDW sorting and recycling processes.

## Abstract

The sustainable management of Construction and Demolition Wastes (CDWs) represents a crucial challenge for the European Union, considering that this wastes stream constitutes one of the main sources of man-made solid wastes. The implementation of strategies aimed at the recovery and recycling of these materials is essential to reduce the environmental impact of the construction sector and to foster the transition towards a circular economy model. However, one of the main obstacles for effective reuse and/or recycling of CDWs lies in the complexity of their composition, which includes a wide range of materials such as concrete, bricks, ceramics, metals, and wood, not rarely contaminated with harmful substances. In this context, this data article presents a comprehensive material library designed to collect, organise, and make available data from advanced material characterisation analyses based on vision systems data. Specifically, the library focuses on data obtained through two measurement techniques: infrared (IR) thermography and hyperspectral imaging (HSI). These methodologies were selected for their ability to provide complementary information on the chemical composition and physical properties of materials. The material library was developed as part of an in-depth study of CDW from building demolition and renovation operations in several EU countries. The data collection process included the preparation and analysis of representative samples, with the aim of ensuring maximum accuracy and reproducibility of the measurements. The data obtained were standardised and organised in a format compatible with the main statistical analysis and machine learning tools to facilitate their integration into predictive models and decision-making processes. The article describes in detail the library structure, data collection protocols, and practical applications in the fields of waste management and sustainable construction. In addition, the benefits of this resource for the scientific and industrial community are discussed, including the possibility of using the data to develop/fine-tune artificial intelligence (AI) algorithms capable of optimising sorting and recycling processes by recognition and discrimination among different types of CDW material using the aforementioned sensors. The material library represents a significant contribution to addressing the challenges posed by CDW management, promoting a more efficient use of resources and reducing the environmental impact of construction and demolition activities. This extensive database not only facilitates material characterisation and separation but also represents a solid basis for future technological innovation in the construction sector.