Natural Algorithms: Biodegradable Landscape Architecture Inspired By Traditional Patterns

This study investigates the integration of bio-digital technology in landscape design, aiming to address ecological and aesthetic challenges by merging machine learning with biodegradable materials. Centered in Hechuan District, Chongqing, our methodology encompasses data collection, machine learning model development, parametric modeling, structural optimization, and plant material testing. Traditional architectural bat patterns were gathered through field surveys and transformed into contemporary designs using Generative Adversarial Networks (GANs) and diffusion models. Parametric modeling with Karamba3D ensured both aesthetic appeal and structural integrity. Architecture students participated in material experiments, ultimately selecting Metasequoia for its biodegradability and mechanical properties. This research highlights the significance of local environments in sustainable design and demonstrates how traditional culture can be integrated with modern technology to create innovative, eco-friendly landscape architecture solutions.