Imagine your pregnant wife is giving birth. Which hospital is closest? Which hospital will give you the quickest? Which hospital has the best care?
What it does
Waiting in an emergency room is one of the most inconvenient experiences in modern life.
Hospitals are trying their best to inform patients of potential wait times. Unfortunately this information is not readily accessible or user-friendly.
find-ER (beta) is a location-based website that lets users know the potential wait times and the travel time to hospitals. Integrates yelp hospital reviews to help users choose which hospital is best for their needs.
How I built it
- Realtime: MUHC, St. Mary's Hospital (Ontario)
- Scaped using Python (BeautifulSoup, URLLib)
- Historical Montreal Hospital Data: Montreal Gazette
- used data from the web to predict emergency room visits
- created a time-series analysis on realtime emergency room occupancy
- used a linear regression to get estimated wait time
- analysis was done in R
- predictions imported to MySQL
Challenges we ran into
Hospital data is fragmented. Real time information about wait times is sparse.
Text scraping the web is one way to gather this data but also very difficult due to reliance on external servers.
Configuration and integration of MySQL database to web application.
Accomplishments that we're proud of
Creating a database from many different sources.
Creating the forecast model.
Integrating all the pieces together in a useable app.
Creating a mobile mockup in 20 hours.
What's next for find-er
Refining the web app. Optimizing performance.