Packetty

Raspberry Pi Wifi Motion Detectors. #CANSOFCOM
Android Application
Packet Strength

Inspiration

Fueled with the desire to explore the opportunity of using WiFi signal's for motion detection, our team came together from different campuses to breakthrough

What it does

Packetty is an intelligent motion detection system built around everyday WiFi transmitting devices. Combined with Azure and Firebase this application provides a robust security solution with real time image recognition and on demand alerts to ensure that the users always have a watch on their property. The users have access to an android application to be notified and keep track of intruders.

How we built it

The first stage of this project stems from a research paper aiming to use WiFi signals for motion detection. Using two Raspberry Pi's with WiFi adapters, we were able to send and receive beacon packets to keep track of the strength. Due to the disruption of packets and fluctuation of strengths we were able to capture motion.

On setup of this module, pictures of the owners are taken and an Azure neural network is trained(Additional pictures also can be added and trained making the neural network dynamic)

Once a motion is captured, we triggered cameras using Firebase to take snapshots of the individuals. These snapshots were uploaded to the Azure Face Recognition API to identify potential intruders.

Potential intruders' images are uploaded to Firebase and notification are sent to the users using our Android Application. Users can then view the pictures and take further action.

Challenges we ran into

Hardware Available
Lack of Documentation
Combining highly diverse modules
Integrating asynchronous programs
Wifi Packet Strength Detection using mathematical formulas

Accomplishments that we're proud of

We are proud of understanding properties of wifi signals, sending beacon packets, measuring the difference in packets strengths, customizing Azure Neural Networks for facial characteristics and building a real time android application in a span of 36 hours. We are happy that we made a diverse project with many applications in the contemporary society.

What we learned

Microsoft Azure Api
Sending and receiving wifi packets on Raspberry Pi
Maintaining a real time application

What's next for Packetty

Implement Google Home and Amazon Echo Support
Bring support to more smart devices
Better Algorithms to further enhance User Experience and improve motion detection