We were inspired by dystopian science fiction where face scanning is a normal part of everyday life.
What it does
When a customer walks into a store a picture containing their face is sent to a cloud service to be analyzed. This information is then stored on a server to track customer metrics so that a store can better serve their customers.
How we built it
The key piece of hardware we made use of is a Raspberry Pi with a Pi Camera. This Raspberry Pi runs a python script that constantly streams images to an Azure service that checks for faces and returns information about the faces that are present. This information is then sent to an Apache server on an Azure Ubuntu VM and stored in a MySQL database, where it can be displayed through a web portal.
Challenges we ran into
We had a lot of difficulty setting up our hardware. The main issue was setting up our Raspberry Pi over serial to connect to the eduroam network. Our PHP code had some inconsistencies that lead to a lot of troubleshooting. About halfway through the hackathon one of our laptops overheated and we ended up using a live Ubuntu drive checked out from the hardware booth.
Accomplishments that we're proud of
We're proud that we finished everything we set out do within the time limit.
What we learned
We gained a very thorough knowledge of how to correctly set up a wireless connection over serial on a Raspberry Pi, as we spent several hours troubleshooting ours. We also became much more familiar with PHP and MySQL.
What's next for Auto Metrics
We want to improve our metrics page with smarter insights to help store owners maximize revenue and improve customer experience.