Artificial intelligence applied to product design for a circular economy, reuse, and efficient recycling

Abstract

This study investigates the application of artificial intelligence (AI) to product design for the circular economy using a sequential mixed method. The qualitative phase, based on interviews and focus groups, identified a "circularity gap" in current processes and key requirements for AI tools: transparent integration, explainability, and human control. The quantitative phase developed and validated "CircularDesign-AI", a prototype assisting design within CAD software. A controlled experiment (n=30) showed that AI-assisted designs significantly improved disassemblability (+62%), material compatibility (+48%), and reduced estimated carbon footprint (-28%), while also cutting design time by 18%. Users reported high usability (SUS=79.2) and perceived utility. It is concluded that AI can overcome practical barriers by operationalizing circularity principles, but its effectiveness depends on data quality, algorithmic transparency, and integration into broader systemic changes in business models and policies. AI emerges as a valuable co-pilot for designers, not a replacement, catalyzing the transition towards regenerative design