Heatmap of SF Parking Availability
The quest for finding parking in a busy city has never been harder. Do you drive around hoping to find an open spot, or do you finesse your way into a spot and hope you don't get towed? Our team wanted an Airbnb style service for street parking to alleviate the woes of the common driver, and thus InDemandParking was born. We developed InDemandParking to connect people looking for parking spots with people that are about to leave their spot, streamlining the search for parking and giving users more time for that concert or fancy dinner that they planned out.
What it does
Users looking to make a few quick bucks that are about to leave their parking spot can list their spot on the app. Users looking for parking at a specific time mark their destination on a map, pulling up a list of spots posted by others. For a small fee determined by us using a dynamic-pricing model, they can reserve the spot, and a percentage of the fee will go towards the owner that posted the parking spot. With this setup, people save time looking for parking for a small fee, while other users about to leave their parking spots can earn a few bucks at no extra cost to their schedule. In addition to reservations, data collected in the form of search queries and spot listings from users is used to calculate "hotzones" of parking, where there are likely to be lots of available parking. People that don't want to pay to reserve a spot can use these recommendations to drive towards a general area to find parking. Looking for parking has never been easier.
In addition, we have also integrated the Ford API provided to seamlessly transition the app from your phone to the comfort of your Ford personal vehicle. Amazing :o
How we built it
The majority of the backend work consisted of creating a database for users and parking spot listings, determining clusters for available parking spaces, and handling requests from users. We used Java due to its scalability. For data analysis, we used Python and Flask to connect to our machine learning microservice. We also used Spring cuz its kinda nice. Due to the scarcity of pretty languages, we had no choice but to stick with React for our front end work.
Challenges we ran into
React would not cooperate. ML libraries were hard to pick up.
Accomplishments that we're proud of
Being able to build such a comprehensive project in the small timeframe that we were given has left us all with a huge sense of fulfillment. We hope that the app is able to impact the lives of many, whether it be saving them precious time or bringing them a few dollars closer to the three comma club.
What we learned
Frontend is a pain haha...
What's next for InDemandParking
Emerging global markets, generalization for other services such as restaurant reservation or cafe seats, and a move away from React. Also Docusign API integration for that $1000 cash consideration.