Emily had an incident last summer where someone swindled money out of her wallet without her knowing. This inspired the team to design a more secure wallet. To make it easy on the user, we wanted to implement voice commands with Alexa, and on board face recognition. When researching current market presence, all options required either a time consuming locking/unlocking mechanism, or were very expensive.
What it does
Its primary confirmation method is using a camera on-board to send a picture to the server which verifies the person as the wallet's owner or not. We use open data on theft and burglary to determine which ZIP codes have a high 'Crime Score' and GPS signals to determine which ZIP code you are in and if the Crime Score is high enough to lock by default. You can also use Alexa to set the default of the lock to locked or unlocked.
How we built it
We started off with a 3D printing base to fit the wallet and could house the camera, raspberry pi, batteries, and the servo. We first built a program to operate the servo based on a simple trigger. From there we made the trigger by Alexa while simultaneously having facial recognition as an option. We used opendata from the city of Louisville on crime rates and population to have a weighted crime score system where most recent occurrences of theft and burglary give a higher score. The raspberry pi connects via wifi to the DragonBoard to obtain the information about faces stored on the system, and gives information on crime scores and GPS location of the wallet.
Challenges we ran into
Our major challenge was configuring and having Alexa give the output that we wanted. We also struggled to find a crime score scale that made sense in terms of the danger in low population areas. Integrating all of the systems together also was an issue, and our case is very bulky and not very convenient, but with more work on the project this product would become smaller and easier to use.
Accomplishments that we're proud of
We integrated all of the challenges that Derbyhacks gave us. We used Alexa, Dragonboard, and Domain name into our project. We solved a real world problem, and while it may not be optimal, it is a good start on an idea that could go a long way.
What we learned
We learned a lot about the systems of Raspberry Pi and new technology with Alexa and facial recognition.
What's next for Derby Wallet
the lock needs to be made more secure while also reducing the size of the product, it needs to become more consumer friendly and be able to be applied to more wallet designs.