Inspiration
While preparing for a trip to Disneyland, I watched YouTube videos from frequent visitors giving tips and advice. When looking at the comments, many people complain that Disney is not worth going anymore. A large reason is high congestion and crowding in the parks. Disney theme parks have an increasing overcrowding problem which gets worse annually. Having high-density areas in parks poses issues in terms of safety, visitor experience, and business revenue.
Components
The project proposes a graphical solution to identify park hotspots and predict where crowds will go in the next 30 minutes. Data needs to be streamed in real-time without being invasive to visitors; transactional data, showtimes, monorail schedules, etc. are potential features for algorithms. The final product is an RShiny applet that works on mobile devices to be integrated into the My Disney Experience mobile app.
How I built it
A large part of the work was on research and development. This includes studying graphical models and identifying which algorithms to use for solving the overcrowding problem. Implementation and testing was completed in RStudio.
Challenges I ran into
Adding the graphical models to the geospatial mapping was a little tricky. However, the transition from mathematical algorithms to building an app was a real challenge... I know two women who studied pure math and went straight to software engineering roles after graduation - I look up to them!
Accomplishments that I'm proud of
This is my first time creating a mapping tool, and I think I did a good job.. Also, I didn't see a single math major there. I want to show that mathematicians can be great hackers too since we have the creativity and mindset to tackle complex problems. I hope this project inspires students studying math to solve problems outside the classroom and put their abstract minds to the test.
What's next for Magical-Graphs
Adding more nodes to the Magic Kingdom network
Adding time dependencies (potentially ARIMAX time series forecasting)
Full mobile app functionality
Log in or sign up for Devpost to join the conversation.