HomeSafe

With a smart home comes a smart security system to detect and monitor intruders.

Comment

Inspiration

We're using next-generation technology to upgrade outdated traditional home security systems which utilize unreliable sensors.

What it does

Our cutting-edge technology feeds auditory information from always-connected IoT devices to advanced machine learning algorithms to detect and classify abnormal sounds and inform the user of the results in real-time.

How we built it

Google Assistant parses spoken or typed commands from the user to enable or disable the home security system. Google Actions sends this command to a database hosted on the cloud through Microsoft's Azure services. Devices in the user's home ,using multi-threading on Intel processors, simultaneously poll the database and analyze sounds in the home through IBM's Max Audio Classifer, sending the user an SMS message through Twilio when detecting abnormal sounds.

Challenges we ran into

Interfacing the cloud and clients together, networking between devices, Google Home Mini APIs being limited. Also, the program was taking too long to evaluate the audio data in real time, hence we needed to employ a multi-threaded application to divide the work among polling and classifying data using Intel processors.

Accomplishments that we're proud of

Smooth integration of various platforms through cross communication in a relatively short period of time. Similarly, achieving cohesiveness throughout the team which had different backgrounds and skill sets.

What we learned

Ask questions of more knowledgeable people when you can without fear of looking ignorant. This leads to professional development and stronger networking skills.

What's next for HomeSafe

Integrating it with future IoT devices that deal with vision and motion to create a solid security system capable of providing the user with more accurate live feedback.

Built With

Share this project:

Updates

Submission history