NTU Uses AI to Advance Circular Economy Policy

Researchers from National Taiwan University have developed an AI-assisted workflow combining multimodal GPT technology with system dynamics modeling to accelerate circular economy policy analysis. Using Taiwan’s material flow data from 2013 to 2022, the team built dynamic simulations exploring nine policy scenarios to guide the nation’s 2030 circularity goals.

The innovative approach enables GPT to interpret complex system diagrams, including industrial activities, emissions, and resource flows, transforming them into executable simulation models. A six-stage AI process (GPT1–GPT6) allows comprehensive evaluation of policy impacts across Taiwan’s material flow network.

Simulation results suggest that Taiwan’s national circular material use rate could rise from 22% in 2022 to 29% by 2030, with resource productivity increasing from NT$65 to NT$88 per kilogram. Prof. Hwong-wen Ma emphasized that this scalable AI workflow provides data-driven insights for policymakers in Taiwan and other regions seeking to implement effective circular economy strategies.