Community shared urban gardens are collaborative spaces where individuals from a community come together to cultivate and maintain gardens within urban areas.
These gardens provide a shared space for growing fruits, vegetables, herbs, and flowers. These can be with fully shared plots and produce or subdivided into individual growing plots.
Building an AI tool capable of:
We implemented the Linear Problem in a Constraint Satisfaction problem (CSP), using the Python library PuLP.
We hypothesized that each city has a certain number of available plots in different random locations and of different random sizes.
We divided the city population into neighborhoods, each one having a randomly chosen size within a certain range and location.
We aim to produce at least 5% of the yearly quantity of vegetables consumed by each community from their assigned plots of land
The total production of each plot depends on the city location/climate by the choice of different vegetables to grow.
Some of the constraints considered for the Linear Problem are introduced here below.
Each neighborhood is associated with only one plot:
Following, the objective function
For more details visit the GitHub repository.
The demand of all connected neighborhoods must be met:
Where i is the iterator for the neighborhoods; j is the one for the plots; x_ij is equal to 1 if and only if the neighborhood i is connected to the plot j; r_i is the demand regarding neighborhood i and offer_j id the production of the plot j (we considered monthly average production).
Choose the best combination of vegetables from a pool of 20. Chooses based on weather compatibility and average yield per squared meter required.
| Plants | Months landfill | Months landfill 2 | Production Time (weeks) | Min Temperature | Max Temperature | Yield per SQM | |
|---|---|---|---|---|---|---|---|
| 1 | GARLIC | October-February | - | 16 | -5 | 25 | 1.0 |
| 2 | BROCCOLI | March-April | - | 16 | -5 | 35 | 1.0 |
| 3 | CARROTS | February-September | - | 10 | 15 | 35 | 4.0 |
| 4 | KALE | April-September | - | 8 | 5 | 35 | 1.0 |
| 5 | ONIONS | October-November | March-May | 12 | 0 | 30 | 3.0 |
| 6 | GREEN BEANS | January-December | - | 7 | 10 | 35 | 3.0 |
| 7 | CHICORY | February-March | July-September | 8 | 5 | 25 | 3.0 |
| 8 | LETTUCE | April-July | - | 6 | 10 | 25 | 3.0 |
| 9 | EGGPLANT | March-June | - | 9 | 15 | 35 | 4.0 |
| 10 | MELON | March-June | - | 10 | 15 | 40 | 4.0 |
| 11 | POTATOES | February-March | September-October | 16 | 10 | 25 | 4.0 |
| 12 | CHILI PEPPER | April-July | - | 8 | 15 | 40 | 0.5 |
| 13 | BELL PEPPER | March-June | - | 8 | 15 | 40 | 4.0 |
| 14 | PEAS | October-April | - | 8 | 0 | 30 | 1.0 |
| 15 | TOMATOES | March-April | - | 10 | 15 | 35 | 4.0 |
| 16 | RADICCHIO | May-July | - | 8 | -5 | 25 | 3.0 |
| 17 | SPINACH | February-October | - | 8 | 10 | 30 | 2.0 |
| 18 | CABBAGE | June-September | - | 10 | -5 | 25 | 5.0 |
| 19 | PUMPKIN | March-June | - | 15 | 15 | 40 | 13.0 |
| 20 | ZUCCHINI | February-August | - | 6 | 10 | 40 | 3.0 |
Based on the city location in the world we gather statistic information about the weather of the prior year to make a wide monthly prediction about the climate for the following year and choose the most appropriate vegetable to plant.
Each neighborhood is connected to the selected plot. All unused plots are deleted. Only 20 available plots are needed to be used as a urban gardens to satisfy the produce requirement of the community.
Many improvements can be performed, this is just a demo required for a school project. Indeed, only 5 cities have been tested and only 20 kind of plants have been selected. This project should be pretty robusts to the implementation of the tests regarding more cities and any kind of plant.
A neaural network can be implemented too, maybe one which recognizes plant and notice the owner to water it or to trash the plant if it looks bad.. so many options are available.
Thank you for reading my paper, if you have any suggestions or corrections do not hesitate to contact me :).
