We wanted to create a proof-of-concept for a potentially useful device that could be used commercially and at a large scale. We ultimately designed to focus on the agricultural industry as we feel that there's a lot of innovation possible in this space.
What it does
The PowerPlant uses sensors to detect whether a plant is receiving enough water. If it's not, then it sends a signal to water the plant. While our proof of concept doesn't actually receive the signal to pour water (we quite like having working laptops), it would be extremely easy to enable this feature.
All data detected by the sensor is sent to a webserver, where users can view the current and historical data from the sensors. The user is also told whether the plant is currently being automatically watered.
How I built it
Challenges I ran into
After choosing our hardware, we discovered that MLH didn't have an adapter to connect it to a network. This meant we had to work around this issue by writing text files directly to the server using Gobetwino. This was an imperfect solution that caused some other problems, but it worked well enough to make a demoable product.
We also had quite a lot of problems with Chart.js. There's some undocumented quirks to it that we had to deal with - for example, data isn't plotted on the chart unless a label for it is set.
Accomplishments that I'm proud of
For most of us, this was the first time we'd ever created a hardware hack (and competed in a hackathon in general), so managing to create something demoable is amazing. One of our team members even managed to learn the basics of web development from scratch.
What I learned
As a team we learned a lot this weekend - everything from how to make hardware communicate with software, the basics of developing with Arduino and how to use the Charts.js library. Two of our team member's first language isn't English, so managing to achieve this is incredible.
What's next for PowerPlant
We think that the technology used in this prototype could have great real world applications. It's almost certainly possible to build a more stable self-contained unit that could be used commercially.