When driving to a location, wouldn't be magical if your favorite navigation app could not only show you parking lot locations, but even individual parking lot spaces available? What if you could be reminded by text that your time was up and add time to that reservation by text as well? We wanted to make a computer-vision powered web-app/REST API parking lot owners could set up easily to attract more costumers and make finding parking in our city breeze.
What it does
park.AI sets up conveniently into your existing security cameras and with a quick setup is ready to broadcast to the world and yourself how many parking spots are available in real-time. It can also accept payments for parking, alert users when their reservation is ending and alert the owner that someone has not left their spot yet when their reservation expires.
How we built it
We built park.AI with a Node.js backend using Express as a router, Bootstrap and jQuery on the frontend and SightHound as our computer vision API. Anything needed to be stored is on a MongoDB.
Challenges we ran into
HTTP request payloads were not being correctly transported and data was unusable at endpoints.
Accomplishments that we're proud of
Integration of many features making our final prototype rich in user-experience goodness.
What we learned
Callback hell in JS
What's next for park.AI
Adding actual card payment processing and Google Maps integration.
Team name: Dream team Team members: Nicolas MacBeth, Jason Botzas-Coluni, Dina Khalesi and Laetitia Ségaux