Bike Sharing Data for Urban Planning

Note: This post was originally published by Luis Natera on his personal blog. It has been republished here as part of TYN Studio's content.

Bike-sharing data provides valuable insights for urban planning. By examining Guadalajara's MiBici bike-sharing system using January 2019 data comprising 405,169 trips, we can understand user movement patterns and identify where infrastructure improvements would have the most impact.

Methodology

I investigated user movement patterns by constructing origin-destination networks with stations as nodes and trips as weighted links. To visualize actual route usage, I employed OSMnx to retrieve street data and applied Dijkstra's algorithm to calculate shortest-path routes for all trips, treating this as a proxy for actual bike commute patterns.

While cyclists don't always take the shortest path, this approach gives us a reasonable approximation of which streets are likely being used for bike-sharing trips.

Daily Network Evolution

This animation shows how the network evolves throughout the month, with station sizes representing usage and the connections showing inter-station trip volumes. You can see daily patterns emerge—busier weekdays, quieter weekends, and peak hours during commute times.

Street-Level Analysis

By overlaying the bike trips onto the actual street network, we can see which routes are most heavily used. This is crucial information for planning where to add dedicated bike lanes or improve existing infrastructure.

Key Findings

The aggregated street usage map highlights primary biking corridors in red. The analysis revealed several important patterns:

• Centralized structure: The city's central area dominates activity due to higher station density
• Hub-and-spoke pattern: Trip patterns reveal peripheral zones primarily connecting to the downtown core
• Limited inter-peripheral usage: Very few trips connect different peripheral neighborhoods

Implications for Planning

This data suggests that expanding the bike-sharing system should focus on:

1. Adding more stations in peripheral areas to increase coverage
2. Creating better connections between peripheral neighborhoods
3. Identifying and improving the most heavily-used corridors with dedicated infrastructure

The heavily-used routes shown in red are prime candidates for protected bike lanes, as they already see significant cycling traffic and would benefit most from safety improvements.

Code available on GitHub: DataVisualization