As school students, we set out to solve an important and challenging task for administrators - keeping their students safe in a large building. One of the most important things to achieve this is knowing when problems happen - fast. We developed a sensor device that can alert school administrators of potential dangers in real-time so they could take the proper precautions for a safe school environment.
What it does
WiFire is an internet of things (IoT) detector that is able to detect smoke, combustible gas leaks, flames, and monitors changes in temperature and sound. WiFire's device is meant to be placed in multiple areas around the school - and all of these devices monitors dangers and sends data up to a secure cloud. WiFire's app alerts administrators to the potential danger in their locations with their very own mobile phone, allowing them to monitor their school in the palm of their hand.
How we built it
We identified multiple dangers, and the sensors needed to monitor these dangers. This was then wired up to a Raspberry Pi, which polls for data. This then feeds the data from the sensors up into the master cloud hosted engine we wrote to handle multiple sensors over Wifi. Our iOS app organizes the devices, and reads their data into a realtime graphical display.
Challenges we ran into
Billy: The webcam microphone was a challenge to get working with the raspberry pi. There was a lot of undocumented code to sift through and a lot of config files to update. Also working with analog sensors on a raspberry pi is a challenge as is has no analog input pins and the signal must be converted first.
Yu: The web service is getting data from Raspberry Pi, processing data, saving data locally in database, and response different request from the mobile app. It's difficult to communicate between these processes and also it's my first time designing a database and using a database in project. The solution we came up was to keep everything in a stream to make the processing efficient.
Michael: Charting historical and live data - it needed a lot of tough thinking to see how the historical data and live data would mesh together seamlessly on the front end. There were a lot of ways we could have displayed the data, but the current configuration was chosen to be the most intuitive and useful.
Accomplishments that we're proud of
We are very proud of WiFire. We've created a sensor array that feeds data in realtime used to alert administrators of danger in critical locations. We didn't make it solely for the hackathon, but took care to create WiFire like a consumer product, so that it could be scaled to multiple devices seamlessly. We've not only solved the problem that we initially set out to conquer, but showed that we could do within a beautiful interface.
What we learned
WiFire was an incredible learning experience for all of our team. On a complex project requiring many different skills, all of us were put out of our comfort zones to get the work done. We hadn't been organized prior to the event, but were still able to develop as a team effectively.
What's next for WiFire
WiFire needs to upgrade the sensors used and find a microphone which can handle high amounts of ambient noise. The data used by WiFire represents a great opportunity for data analysis. It may be possible to predict the direction of a moving threat based on movement through multiple sensor fields. WiFire could benefit from a comprehensive online portal that could help outside parties, such as the police, make informed decisions in emergency situations.