This study presents an innovative planting design methodology based on the use of Cellular Automata (CA), with the aim of incorporating interactions between plant species into the design process. Unlike traditional methods, which generally distribute plants according to primarily formal or compositional criteria, this approach considers competition, coexistence, and development among different individuals as determining factors in the organization of planted space.
The objective is therefore to incorporate into planting design a factor that is often difficult to predict: the interaction between different plant species and the way they compete, coexist, and evolve over time.
The methodology developed, which we call Natural Neighbouring Behaviour Cellular Automata (NNB-CA), is inspired by John Conway’s well-known “Game of Life.”
The method is structured in 2 phases:
- – Phase 1: site conditions —solar exposure, wind, topography, and soil characteristics— are assessed to determine which species can occupy each area of the project. The NNB-CA is based on the idea that the position of a plant depends not only on the environmental conditions of the site —such as orientation, solar radiation, wind, topography, or soil characteristics— but also on the species surrounding it. For this reason, the system analyzes the site, divides it into a grid of cells, and determines which species are suitable for each cell.
- – Phase 2: the algorithm successively applies a series of rules related to the behavior of each species, including growth rate and form, height, horizontal development, and its ability to compete with other plants. Project-specific criteria can also be incorporated, such as limiting certain groupings, placing aromatic plants close to circulation routes, or avoiding plant combinations that are incompatible from an aesthetic perspective. The process is iterative: each plant interacts with its neighbours, and the distribution changes generation after generation until a stable configuration is reached. The result is a planting proposal that combines site adaptation, species compatibility, and functional and landscape design criteria.
To validate the system, we compare the behavior of a digital model with a real experiment using 3 species of fast-growing aquatic plants. The results show that the digitally generated patterns are similar to those observed in the natural process, demonstrating that competitive relationships between species can be encoded through simple rules and reproduced computationally.
The methodology is also applied to a landscape design case involving pre-existing vegetation, where the algorithm makes it possible to define the introduction of new species while respecting existing plants and the competitive relationships among them.
Conclusion and practical application
This approach makes it possible to generate planting schemes that are better adapted to site conditions, reduce maintenance requirements, and systematically incorporate ecological criteria. Planting design is conceived as a dynamic system rather than simply as a static composition. In addition, the methodology opens up new possibilities for the development of digital tools capable of simulating plant ecosystems and supporting more sustainable and resilient landscape design projects.
For further information about this project, see the following publication:
Bilurbina, M., Abellan, S., & Christodoulou, M. (2019). Landscape Design Methodology: Pattern Formation Through the Use of Cellular Automata. Journal of Digital Landscape Architecture, 4, 95-104.