The challenges + How do we go about making restaurant pick ups faster and safer?
Reading the code will show you more than our terrible presentation. Reading this Devpost is a second best.
What it does
This application uses OCR technology to pick up license plate numbers when a customer arrives to a restaurant. Our OCR API, Cloudmersive, is able to recognize license plates with images and videos. We leveraged this technology to stream our camera feed in order to recognize when customers arrive at the location of the vendor while also using proximity algorithms such as Levenshtein Distance to account for the inaccuracies of the OCR Technologies. The vendor side of our application has the ability to update the status of a customer's order with a simple drag and drop interface and when the customer arrives at the curbside of their location. The system would notify the vendor that the customer has arrived.
Our server is able to not only provide the vendor with the information from the client based on their license plate, but they also have the ability to create new menus, take online orders, and improve traffic to their stores. Within our application we are leveraging the Capital One API to be able to create purchases seemingly within the app so that the Vendor does not have to deal with their own Payment Systems.
On the client side of the application, the customer can sign up for an account with their license plate, that way when the customer arrives, their order is delivered straight to their car with minimal contact. The customer can choose from many restaurants and checkout their cart from the restaurant's menu and track their order status, so they know when their order is ready to pick up. They can also keep track of the orders from within the application so that they know when it is time to get their food quickly and safely.
How I built it
The Frontend was built using React and fully custom styles and animations using Styled Components. The backend was built using Node.js, Express.js for the web server leveraging the Capital One API to handle the purchase transactions, Cloudmersive to handle the OCR needed to read a license plate, MongoDB Atlas, and Mongoose.
Challenges I ran into
OCR not picking up the proper letters/numbers + ratio aspect issues.
Accomplishments that I'm proud of
The amount of work we achieved over the past weekend
What I learned
Way too much to list
What's next for congenial-pancake