How do we populate a room with different plants?
1. Introduction
Our design question is rather simple — how do we populate a room with different types of plants? In our scenario, our aim is to create a room filled with plants that have the highest possible total value while maintaining a good diversity among them. We assume that we will have three different types of plants inside our room. If a, b, and c each indicates the amount of plant A, plant B, and plant C, while, x, y, and z, each indicates the individual price of plant A, plant B, and plant c, our optimization goal will be maxiumizing the value of all plants within the room = ax + by + cz. In order to keep plants diverse, we will also need to maxiumizing plant diversity which can be represented as the reciprocal number of the standard deviation among a, b, and c.
2. Data input
In order to solve this problem through generative design, we need three types of data
2.1. Desired total amount
This parameter will be determined at the beginning. Our algorithm will try to get the final total amount close to this number input. However, due to the method in which the script is written, the computed final total amount will always be smaller than this number input.
2.2. Basic parameters of each plant type
2.3. Daylight analysis of the chosen room — in our case, Avery 600S
3. Behavior logics
Plants need daylight, obviously. Therefore, clustering towards desired daylit areas is the first behavior we implemented for all the plants. Each type of plant has its own desired level of luminance, and therefore will be drawn to regions that receive sufficient daylight for that kind of plant.
Also, we did not forget that each plant creates a shadow radius with their canopy. Naturally, the second behavior will be collision that asks plants that are placed too close to move away from each other.
4. Methodology
This project explores how to maximize the amount and the total value of plants in a given space while maintaining the diversity of the types of plants. The whole process begins with the sunlight analysis of the digital model through Climate Studio and the identification of three primary types of plants. These two major sources of information then generate four parameters for optimization.
Surface — The boundary of the room defines where the plants can populate randomly.
Point Cloud — Points with high value of daylight attract plants to desired places.
Center Points — The center points determine the location of plants.
Radius — The radius of the shadow casted by plants themselves helps to keep plants away from each other in case of blocking the sunlight.
With the parametrics, randomly-populated plants are firstly drawn to windows via the grasshopper component ‘Closet Points’. Then the python algorithm ‘collision’ is implemented to prevent the overlapping of the shadows by themselves. After culling plants falling outside the room and sufficiently lit regions, the final product is passed to Discover for optimization.
5. Result
After running more than 20 generations with each generation testing out 10 different designs, the program produces two extreme designs — one represents the most diverse option while the other indicates the highest value option. In order to fulfill our design goal — generating the design with the highest possible total value while maintaining a good diversity among them, we decide to pick the design option that is positioned right in the middle between these two extremes. We believe this option adapts to the real market that appreciates both value and diversity at the same time and therefore we see it as the most optimal and balanced design.
6. Conclusion
Our design model optimizes total plant value while maintaining a level of diversity within the studied space. The model will offer a range of design options that swing between having the largest possible value and highest diversity determined by the standard deviation of the amount of different plant types. People will be able to choose the best option according to their priorities — higher value or better diversity. The end design option also provides an optimal layout for each plant’s placement according to the daylight condition within the room. The overall design provides some useful answers to the question — how do we populate a room with different plants?
