Matrix-Based Approach to Optimizing Disassembly Sequence Planning for Aluminum-Framed Windows

The construction industry is under pressure to minimize its environmental impact, especially in environments where resources are limited and ecological concerns are increasing. Disassembly has emerged as a crucial strategy for facilitating material recovery at the end-of-life (EoL) stage.

However, the disassembly process remains complex and necessitates strategic planning to optimize material recovery and reduce waste. Unlike most products, windows are larger, more intricate, and composed of multiple materials, layers, and subsystems. This complexity requires comprehensive disassembly planning. Graph-based planning often struggles to accurately capture the extensive range of precedence relations present in complex, multi-material products.

To address this limitation, matrix-based representations have been introduced to convert graphical data into computable formats. Matrices can effectively represent geometric, material, and technical precedence relations embedded in 2D cross-section drawings, allowing for the transformation of this information into matrix data that algorithms can use to generate disassembly sequences.

This study reviews both adjacency and incidence matrices, employing unweighted and weighted data to develop disassembly sequences. It emphasizes the capabilities of these models in capturing the geometric and material-based relations among aluminum-framed window parts within the parametric modeling environment. Additionally, it identifies key factors, dependencies, and challenges associated with the matrix-based disassembly of multi-material window systems.