AutoArmor | Devpost

Code for PIR sensor

Inspiration

Vehicle thefts have significantly increased in the United States. According to the National Insurance Crime Bureau, around 1,020,729 vehicles are stolen in 2023. Unfortunately, it is known that police solve only a fraction of the cases. The CBS News analyzed the police data and determined that 9 in 10 car thefts are unsolved. Unfortunately, my Kia Soul was broken into, and this sparked me to want to create an anti-vehicle theft device.

What it does

To prevent and find the whereabouts of stolen vehicles, AutoArmor can be installed. AutoArmor is a device that uses a PIR infrared motion sensor to detect activity around the vehicle and it comes with a camera that is installed on the car's dashboard to record the activity detected by the PIR sensor (the camera can spin 360 degrees; it turns towards the direction that motion is detected). PIR motion sensors are designed to detect the radiation emitted by living beings. AutoArmor is also connected to an app on the phone which enables more options for security measures for the user. Through the app, users can set a time during the day when they might want camera footage when the PIR motion sensor detects someone near the car. Most car thefts are bound to happen late at night or early in the morning, so this feature enables the user to have footage of the suspicious activities that happen near their car. Through the implementation of AI, patterns of the activity around the vehicle can be analyzed to distinguish between the normal movements (such as someone walking by) and suspicious behaviors (such as someone trying to break in). This analysis performed by AI can help prevent false alarms. AI algorithms can learn the typical behavior around the vehicle and flag any deviation from the norm. For example, if Auto Armor detects activity during unusual hours or in unexpected locations, it can alert the user through their phone.

How I built it

Through Arduino, I used a PIR motion sensor used to detect human activity around the vehicle. The PIR (passive infrared sensor) is an electronic component which senses motion using a pair of pyroelectric sensors to detect heat energy in the surroundings environment. For the microcontroller board, I used an Arduino Mega 2560 board. To test the efficiency of the PIR sensor and the amount of time it takes for the sensor to identify activity, I used a LED light as an indicator of activity detected. In theory, I would make sure activity around the vehicle triggers a camera to start recording the activity. The camera would be connected to an app I created using Figma. This app uses AI to identify the action performed around the vehicle and it notifies the user regarding the activity with an outline of what activity happened around the user’s vehicle.

Challenges I ran into

I ran into a few challenges while programming the PIR sensor. First, I didn’t know from what distance the PIR motion sensor should detect activity at. I had to contemplate scenarios that could happen. Let’s say a person happened to be causally walking past a car with no malicious intent, the camera shouldn’t be triggered, but activity of a person walking can be noted in the app. However, if someone were to touch the car for more than 20 seconds, the camera should start recording and an immediate notification should be sent to the app. So, I devised that 1ft away from the car is an appropriate range where the PIR sensor should start recording the activity. But then I got an idea, people have different preferences so I set the 1ft as the standard distance and through the app people can change the distance, they want their PIR sensor to detect activity from.

Accomplishments that I'm proud of

Something I am really proud of is how I was able to create a possible solution to a problem faced by many people, including myself. While this solution might not guarantee the security of the vehicle, my intention in creating this idea was to ensure that vehicle whereabouts and security are right in the palm of your hands (in your phone). While I wasn’t able to code the app and implement AI in real time, the prototype is a good baseline for the future when I can implement and create the app.

What I learned

I learned about my creative capabilities. I took a long time to think of this device, the logo, the design, the application, etc. I am proud of this application, and I hope I can work to bring this device to reality in the future:).

What's next for AutoArmor

People have to download the AutoArmor app, so they get notifications on suspicious activity detected around their car. Through the app, the user can change and adjust a few settings, like the distance the PIR sensor can detect from. Through AI, the video recorded from the camera can be analyzed to distinguish between the normal movements (such as someone walking by) and suspicious behaviors (such as someone trying to break in).

I also want the camera on the dashboard to have decor people can put on it to disguise it.