Inspiration
1 in 2 women feel unsafe at night, compared to the 1 in 5 men. This is not only a statistic, but a limitation on their freedom and well-being.
Coming from a family of 2 older sisters and an overprotective dad, this project was something that was very important to me and was an excellent opportunity to use Google's ai to make a solution for this problem
What it does
The idea of the design is a hat that alerts the user of any nearby threats using omnidirectional cameras. Thanks to AI, this design stands out as it not only helps you be more aware of their surroundings, but also takes action automatically in the case a person does actually attack. Some actions can include contacting the authorities, contacting loved ones, and making loud sounds, alerting anyone nearby of the danger.
As of now in the code, the device has three modes, red, orange, and green. Green means that the person is within a safe distance and nothing happens. Orange means someone is close, however not too close, this mode can be treated like a buffer before immediately taking drastic action. Red means the person is extremely close, which will make the device take more drastic measures to help the user.
How we built it
This project was mainly built using Ai. To be more specific, this project uses Google Cloud Vision API for accessing the camera, and uses the Ai to help detect any individuals. Python Code is then used to take actions depending on the proximity of the person.
Challenges we ran into
As Alice Keeler once said "Try something you know nothing about". So for this hackathon, I thought it would be a good challenge to try gender equality. At first, I had to think hard about what I took for granted on what women are limited to, which sparked the inspiration for this project
Another problem was accessibility, a huge problem to tackle was about how to use the design. At first, I had originally came up with an app that would record the user's surroundings, however, I realized that a phone was unreliable for an omnidirectional view. So I had to think about how else someone could easily use the cameras, experimenting with several designs such as the use of wristbands or leg bracers. Eventually I decided to go with a hat due to it's ease of use, and it could utilize the hat's elevation for better scanning.
Accomplishments that we're proud of
Something I am proud of is how I was able to use AI to help build a project that started from an idea, to a software application that has real world impact.
What we learned
What I mainly learned from this is that there is a lot more uses of Ai than just a LLM. Ai can be used to help build projects, and easily help spark ideas one might have to an actual reality. Although this code was mainly Ai generated, I did learn a good amount of Python via reviewing the code and understanding how the program works.
What's next for Scanner Sweep
Hardware: In terms of hardware, there a few next steps to include. The main thing we want to explore is making the hat feel comfortable to wear, while also balancing camera view, battery life, and weight of the equipment. Another feature to explore is being able to detect sound, in the event the cameras do not detect any threats, the device should be able to detect the user's screams and take immediate action, even if no threats were detected nearby.
Software: The next step in terms of software for this is creating an app that allows for easy customization of the hat's settings. This will allow users to change certain features, sensitivity, and different. We understand that each user is different, and the idea of the app will help match those user's personal preferences.
Log in or sign up for Devpost to join the conversation.