A team of researchers from Monash University and Charles Darwin University (CDU) have developed a new artificial intelligence (AI) system that is capable of automatically identifying contaminated construction and demolition wood waste.

The team developed what it suggests is the first real-world image dataset of contaminated wood waste, which promises to potentially lead to smarter recycling and sustainable construction.

Dataset samples. Source: Resources, Conservation and Recycling, DOI: 10.1016/j.resconrec.2025.108278

To accomplish this, the team trained and tested deep learning models to detect contamination types in wood waste via images. The team opted to do this because contaminated wood from construction and demolition sites typically ends up in the landfill due to the challenge of sorting it manually. However, when applying AI models, the team identified strong precision and recall across six different types of wood contamination.

"We curated the first real-world image dataset of contaminated construction and demolition wood waste. This new system could be deployed via camera-enabled sorting lines, drones or handheld tools to support on-site decision-making," the team explained.

The project demonstrated that by refining advanced deep learning models — such as convolutional neural networks (CNNs) and transformers — these systems could automatically identify types of contamination in wood using standard RGB images.

The researchers believe that this system might one day support wood waste reuse, recycling and reclamation, serving as a tool for recovering valuable resources and reducing landfill dependency.

The findings are detailed in the article, “Automated recognition of contaminated construction and demolition wood waste using deep learning,” which appears in the journal Resources, Conservation and Recycling