A Constraint-Driven Framework for AIGC-Assisted Landscape Scheme Development and Optimisation

Authors

  • Xiaobin Li Liaoning Communication University, Shenyang 110000, China

DOI:

https://doi.org/10.6981/FEM.202612_7(12).0022

Keywords:

Generative AI, landscape architecture, constraint-driven design, scheme reasoning, human–AI collaboration.

Abstract

With the rapid iteration of AI technologies, their application in landscape architecture has become increasingly common, shifting design practice from experience-based manual drawing toward a collaborative model of data-assisted generation and value judgment. However, scheme reasoning in landscape architecture must consider multiple constraints, including codes and standards, site conditions, ecology, engineering, cost, and maintenance. Current AI tools cannot adequately handle these constraints; their outputs are often visually usable but difficult to implement. This paper takes constraints as the organizing logic and proposes a “three-tier constraints–three-stage–dual-loop” human–AI collaborative framework for scheme reasoning. The framework consists of constraint coding in the early stage, generative scheme reasoning in the middle stage, and scheme evaluation and iteration in the late stage, supported by a dual-loop mechanism combining AI iteration and human decision-making. An exploratory practice involving a residential children’s play area preliminarily indicates that constraint coding can improve the usability and evaluability of AI outputs. Generative AI is suitable for rapid multi-style variant generation and imagery reasoning under constrained conditions, whereas engineering responsibility and design direction judgment should remain with designers.

Downloads

Download data is not yet available.

References

[1] Wang, S. J., Tian, J. W., Wang, G. Q., et al. (2026). Site constraint transfer and multi-model collaboration in auxiliary generation of rural landscape plans. Landscape Architecture. https://doi.org/10.3724/j.fjyl.LA20260407. (in Chinese)

[2] Wei, C., Du, Y., Huang, X., et al. (2026). A cross-stage human–GenAI collaborative workflow for landscape architecture design: a pocket park feasibility study. Buildings, 16(19), 3819. https://doi.org/10.3390/buildings16193819

[3] Zhu, Z. B., Song, Y. F., Yue, B. R., et al. (2026). Three roles of AI in knowledge production in landscape planning and design practice: servant, opponent, and partner. Chinese Landscape Architecture, 2026(3). (in Chinese)

[4] Generative AI-driven architectural and structural engineering designs: a review. (2026). In IEEE.

[5] Shao, Y. H., Lin, P. Y., & Liu, S. (2026). Context modeling and application boundaries of generative artificial intelligence in landscape architecture design assistance. Journal of Chinese Urban Forestry, 24(2), 103–109. (in Chinese)

Downloads

Published

2026-11-16

Issue

Section

Articles

How to Cite

Li, X. (2026). A Constraint-Driven Framework for AIGC-Assisted Landscape Scheme Development and Optimisation. Frontiers in Economics and Management, 7(12), 165-170. https://doi.org/10.6981/FEM.202612_7(12).0022