Improve management & efficiency of bicycle sharing business in Washington, D.C.
- Summarize spatial movement of fleet in an informational map
- Define high and low demand times
- Improve fleet management (e.g. placement of bikes or planning maintenance) based on spatial demand & usage
- Predict user type (registered vs. casual) based on time of day or week
The client is Capital Bikeshare, and will be able to understand spatial movement and usage patterns of their fleet, and improve efficiency by ensuring bikes are where they need to be when in demand, and maintenance is occurring at the most opportune times (preventing dissatisfied customers).
Capital Bikeshare posts quarterly data reports of bike trip times, start and end locations, and type of user (registered or casual). Each trip is on one line of data. These data are readily and publicly available at https://www.capitalbikeshare.com/system-data.
I intend to visualize the data using mapping techniques, and hope to develop some schematics to represent fleet movement. Histograms indicating high and low demand times at the various bike storage locations will help to develop more thorough analysis such as predicting user type based on time of use.
The final products will include code generalized for use with any quarterly report of bike usage data, as well as a slide presentation and blog post to describe the methods and benefits to Capital Bikeshare.