Fab Academy 10

Applications / Implications

What If...

many farmers in the world could test
how the plants they grow nowadays
would behave in other more drastic climates,
and therefore create a network of knowledge
that would help them all to be more resilient
to climate change?


The climate is predicted to change radically in the next 20 years, and the future climates of some cities will be more similar to the current ones of others, also generating floods, droughts, drastic increases and decreases in temperature, in places where they did not have before. 

This will affect what farmers will be able to grow and how they grow it. Therefore, I think we will need ways to understand how plants would behave when the temperature rises or falls, or when it has less or more force and time in daylight, as well as excess or lack of water.

ESP8266 12-E Chip Pinout


Plant Observatory
of Weather Adaptability for Resilience

P.O.W.A.R. is a machine that can replicate different climatic conditions like sunlight times, temperature and pluviometry from future climate forecasts, to analyze which are the optimal conditions to grow each plant in the future or a certain place.

This will work as a plant observatory, in which 3 different climate simulations will be compared, changing only one variable at a time, to detect which plant grows better than the other and with this, it will help farmers generate a database to optimize what they will grow in the future.

The future of P.O.W.A.R. is that it will learn by itself through machine learning and image recognition, feeding a global database that helps analyze plants growing in different weathers.

The variables that can be changed inside are:

- Temperature (Cold and Hot)

- Humidity / Rain (Amount of Water)

- Sunlight (Amount of Sunlight and Times of Light)

Sources of Inspiration

Similar proyects or technologies


Arduino Indoor Garden

In this project they created their own board using an Arduino Nano and a bluetooth usb, some sensors and actuators, but he even designed tha housing for everything, to take care of a plant.


Simple Weather Station with Amazing Web Monitoring

This projects shows how to turn your weather data into a really nice graphic interface that you can check from the web.


Online Weather Station (NodeMCU)

In this project they webcrawl data from the Open Weather Map API, and show it into a LCD screen.


The Weather Report

This is Felipe Santos  Fab Academy 2019 artistic project that simulates some weather conditions inside a box.


Lorem ipsum dolor sit amet, consectetur adipisicing elit



4 - Transparent acrylic boards 40x70 cm
6 - 40x20 Pywood Boards (top bottom/box)
3 - 40x40 Plywood Boards (Bottom/Top electronics box)
1 - Plastic Organizer Box
1 - 6mm nylon/silicon tube hose (3mt)
2 - 3D Printed Hinges (+ bolts and knotts)
1 - 3D Printed U Shaped Door Handle (+ bolts and knotts)



1 - Moisture Capacitive Sensor
1 - Temperature Sensor



1 - DC Water Pump
1 - Growing Lights for plant (To Be Defined)
2 - 120x120 mm Fan
1 - LCD Screen


Other Electronics to define:

Power Source



(5mm thick acrylic panels)

- 1 Base (600mm x 600mm).
- 2 Lateral (600mm x 1200mm).
- 2 front and rear (590mm x 1200mm). 



(10mm thick wood panels):

- 1 Base (620mm x 620mm).
- 2 Laterales (600mm x 200mm).
- 2 frontales y traseros (620mm x 200mm). 

(10mm thick wood panels):

- 1 Base (620mm x 620mm).
- 2 Laterales (600mm x 200mm).
- 2 frontales y traseros (620mm x 200mm). 

The top and the bottom part are exactly the same, but what changes is the possition and orientation.



(5mm thick acrylic panel):

- 1 Base (590mm x 590mm). 

HOW IT WORKS? (how is it made)

It´s an acrylic panel with laser cutted holes of 5mm, with a 30 mm separation between them for the water to pass by, into the bottom tank (plastic organizer cube).

Depending of the plastic organizer box that I get, I couls use the top part of it as filter, by drilling or laser cutting some holes, deppending on the matterial.




This would be the system that pumps tha water from the tank up, and drain it again above the plant. I was thinking that I could either use the tubes to generate a net in the upper part whit some holes for the wather to flow, or replicate Felipe Santos System, but I don´t really need them to be randomly assorted, but I would like to have them coverrong all other or with equal spreading. Another option could be to use a "3D printed shower head", which could make the same effect of spreading.

This ones could be made with a 6mm Nylon or Silicon tube hose, with some wholes in the upper part.



Tasks Timeline

This is the order I think I need to follow, to make my project.


Circuit Building

Design and Mill the Development board
(in case I don´t use a NodeMCU).
Connect and test inputs and outputs.
Write the Code.

Machine Box

Design and Build

Design and laser cut the Acrylics
Design and Laser cut the Plywood.
3D Print the Handle, Hinges, bolts and knots.


Web Crawling

Connect the machine to a weather Api where it could take the information to do the simulations, or to a website where you could decide which conditions you want to have.

Current state of the Project

I see this project bigger than I described it, so in order to take it to where I want, what first evaluate some steps and then start evaluating others in the future.

  1. My first goal is to be able to replicate a climate condition inside, including: A certain amount of light in different, a certain amount of wather and a certain temperature in different moments of the day. (replicate a day inside).
  2. For my second goal, I should be able to replicate a whole week with different variables of climate conditions. For explample: A sunny day, a rainny day, a foggi day, a day with sun in the morning and rain at night, a not so sunny day, etc.
  3. My third goal would be to web crawl information from a source, like a weather API to simulate a weather from around the world or a webpage I design where you can progam the simulation wanted.
  4. My fourth goal would be to grow something in the inside of the machine, that could be grown in some other city or weather different than the local one (Barcelona) inside of the box.
  5. **My fifth goal would be to grow 6 exact same plants, in 3 weather conditions and visually (human-made) check their differences and start training a model to predict the best coditions for that plant.
  6. **My sixth goal would be to add a cámera to the machine to start training machine learning models that work with image recognitions, so that in the future the machine tests and learns by itself which is the best way to grow a plant.


(5th Goal)

1. A total of 6 plants will be cultivated, which will be divided into 3 groups of 2 for each set of variables (A1, A2, B1, B2, C1, C2).
2. Each group of plants A and B will have an increase and decrease in one of the variables versus the group of control plants C. (A +, B-, C =)
3. 2 plants are used for each group, to rule out that the results are by chance and the results of the 2 will be averaged to give the result of the group.
4. The height, amount of leaves, size of bigger leaves, amount of fruits and size of fruits will be measured every 3 days.
5. Each test cycle will depend on the harvest times of each plant.
6. The process will be started again by changing a second variable in A and B, but keeping control C the same as the previous experiment.
7. With 9 experiments of 3 groups, the effect of the different variables on plants can be interpreted.


Develop a free web page with Mobirise