While all four members of the team were last minute walk-ins, that did not discourage us from trying to think about something creative, interesting and challenging at the same time. After a productive brainstorming session, we decided that what matters in our lives is security, so why not make something to improve the security of our own homes? That is how we came up with an idea to use motion sensors, camera and an AI to make our homes safer.

Every day thousands of alarms are triggered in fear that someone is breaking in to a house. However, many of those alarms turn out to be false alarms (around 90% according to different statistics right here from Canada). Edmonton police say that 98% of calls from house security systems are false alarms. Not only is the police force wasting resources for these false alarms, but also homeowners need to pay extra fees for false alarms. Earlier this year, Edmonton police proposed an additional $15 fee for houses that install the house security system. The false alarms are raising costs of security services, and are putting a burden on the already busy emergency forces. At the end of the day, this affects every single one of us. Although we can’t eliminate all of the false alarms, we can limit the ones caused by human error. Whether it’s your family member walking in the house without knowing about the alarm, or you yourself forgetting about it, our smart system could mitigate some of these false alarms.

What it does

The program is a unison of different hardware components and software. In short we have rigged a motion sensor which upon trigger sends a signal to a camera to take a video of the area where the sensor has picked up movement. The camera then sends the video to a server, and also notifies the homeowner of an “occurrence”. The homeowner can view the video of the occurrence on their phone in a .mp4 file format.

How we built it

We used the ESP32 module along with the HC-SR04 ultrasonic sensor to detect objects that pass through a window frame. When the ultrasonic sensor detects an object in its range, the ESP32 connects to a nearby smartphone via Bluetooth and sends it a request to begin recording using its camera. The smartphone is aimed at the window frame so that it can record who/what is passing through the window. The smartphone records a 10 second video of what is happening at the window, and sends this video to a web server that stores the videos locally on its database. A second smartphone (which is the one the users use on a day-to-day-basis) connects to this web server to retrieve a list of video recordings, and allows the user to view any of the stored recordings.

Challenges we ran into

This was our first time creating applications that communicated to other devices via Bluetooth and WiFi. Learning how to incorporate these features into our applications took a lot of time and was definitely a steep learning curve for us (but definitely worth it!).

Accomplishments that we're proud of

We are proud that even at 9:00 PM when we were not sure if we can participate in this incredible journey of a hackathon, we quickly rallied and figured out an interesting problem to solve and worked together to make our software work.

What's next for ProTech

Continue the development of the software. While we reached a lot of goals that we set out to do, we did not fulfill all of them. For example, we would like to integrate AI into this system, so that the video that is sent to your phone is analyzed beforehand to see whether the person in the video is a family member or an unknown person (this will help to further reduce the number of false alarms).

Share this project: