In the future we would be driving electric cars. They would have different automatic features, including electronic locks. This system may be vulnerable to the hackers, who would want to unlock the car in public parking lots. So we would like to present CARSNIC, a solution to this problem.
What it does
The device implements a continuous loop, in which the camera is checked in order to detect the theft/car unlocking activity. If there is something suspect, the program iterated in the list of frequencies for the unlocking signal (315MHz for US and 433.92MHz in the rest of the world). If the signal is detected, then the antenna starts to transmit a mirrored signal in order to neutralize the hacker's signal.
We used the propriety that the signal from car keys are sinusoidal, and respects the formula: sin(-x) = -sin(x).
How I built it
We used a Raspberry Pi v3 as SBC, a RPI camera and a RTL-SDR antenna for RX/TX operations. In order to detect the malicious activity and to analyze the plots of the signals, I used python and Custom Vision API from Azure. The admin platform was created using PowerApps and Azure SQL Databases.
Challenges I ran into
The main challenge was that I was not experienced in electronics and learned harder how to work with the components.
Accomplishments that I'm proud of
The main accomplishment was that the MVP was ready for the competition, in order to demonstrate the proposed idea.
What I learned
In this project I learned mostly how to work with hardware embedded systems, This is my first project with Raspberry Pi and RTL-SDR antenna.
What's next for CARSNIC
In the next couple of months, I would like to finish the MVP with all the features in the plan: 3D scanning of the structure, acclimatization and automatic parking from an intelligent service sent directly to your car. Then I think the project should be ready to be presented to investors and accelerators.