For this hackathon, we wanted to solve a big problem: the freshwater crisis! Approximately 60% of our freshwater goes towards agriculture. At the current rate of freshwater consumption, we will not be able to sustain the growing population of humans on Earth. We can make agriculture more efficient to reduce our freshwater consumption by growing crops using hydroponics, which is growing plants in sand, gravel, or liquid without soil.
Benefits of a hydroponics setup:
- Uses up to 67% less water.
- Complete control over nutrient balance
- Higher yield, better results
- Year-round growing
- Space efficient
We were looking at solutions and realized that most of these setups were manually controlled and could be made more efficient by automating a lot of processes. This would benefit busy individuals who want to eat healthy but do not have the time to maintain a garden. Our product lets them grow food with minimal effort.
What it does
Our solution uses machine learning to find the optimal path to maximum yield. It collects data from the nutrient sensors and finds the best combination of factors to achieve the highest output. If a plant is not performing well, it has the ability to adjust the lighting, humidity, temperature, and nutrients it is being given.
The user can check how well the plants are doing on the internet. They can get notifications if any nutrient tank is running low ahead of time. It can also predict when a plant will fruit using historical data, and if the yield will be suboptimal.
How we built it
HydroPlanet incorporates these technologies:
Qualcomm Dragonboard 410c
- Interfaces with sensors and sends data back to database on server
- Does computations necessary for predictions
- C, C++, Python
- Pulls in data from the database and displays it to the end user on a dashboard-style layout.
Challenges we ran into
We never used a Dragonboard before. We weren’t sure how to get the Dragonboard to connect with the sensors and send data back to the website. It took us some time to learn how to use it properly.
Accomplishments that we're proud of
- We are proud that we got the Dragonboard to work and display the data. Same goes for interfacing the sensors with the Arduino and showing it on an LCD.
- The logo
- How the website looks, feels and functions
What we learned
- How to use FTP to make a site go live
- How to set up a domain
- How to use a Qualcomm Dragonboard 410C
- How to use an Arduino Uno
- How to use Linux
- Growth with C, C++, Python
What's next for HydroPlanet
Connecting the Qualcomm Dragonboard with the web server and sending data in real-time
Formatting the sensor data into a csv format for further analyses to be made
Learning Google Cloud Platform
- Google Cloud Machine Learning to handle prediction computations
Building an enclosure with complete sensor components to test the solution and rigorously prove how much more efficient it is compared to a manual hydroponics (the status quo) system