A new AI system for artificial intelligence (AI), which was developed by researchers from Monash University and Charles Darwin University (CDU), has shown an accuracy rate of 91 percent in the identification of contaminated construction and demolition waste.
The study published in the journal Resources, nature conservation and recyclingPresented what the researchers call the first real image data set of contaminated wood waste.
The project was made by Madini de Alwis, a doctoral student at Monash University Department of Civil and Environmental Engineering, and Dr. Milad Bazli from CDU under the supervision of Associate Professor Mehrdad Arashpour, head of the civil engineering system at Monash.
Contaminated wood, which is often spoiled with color, metals, chemicals and other residues, is a major challenge for recycling systems. The difficulty of manually sorting such materials has resulted in a lot of this ending on landfills.
The new AI system, which is operated by Deep Learning models including folding networks (CNNS) and transformers, can automatically recognize six types of contamination in wood using standard -RGB images.
“This new system could be provided via camera-capable sorting lines, drones or hand tools to support the decision-making process on site,” said de Alwis, who curated the image data set as part of research.
The work of the team builds on existing computer vision techniques that are used in general waste flows, but apply a previously lower area, especially to wood waste.
“Due to the fine-tuning, the most modern deep learning models we have shown that these tools can automatically identify the contamination types in wood,” said Dr. Bazli.
“This opens the door to scalable, AI-controlled solutions that support the reuse of wood waste, recycling and recovery.”
According to Monash, Holz is one of the largest components of construction waste worldwide, and a large part of it is sorted properly.
By integrating AI into waste development systems, the technology could significantly reduce the costs and the complexity of processing construction waste.
“This is a practical, scalable solution for a global waste problem,” added de Alwis. “By activating automated sorting, we give recyclers and contractors a powerful instrument to restore valuable resources and reduce the dependency on landfill.”
Doi: 10.1016/J.Resconrec.2025.108278