Rebalancing study

This study focuses on a few of the most intractable problems that any robust bikeshare network faces: rebalancing stations so that they are neither full nor empty, and bike availability. Uneven demand and traffic flows are the root causes, preventing bikeshares from reaching their full potential as forms of public transportation.

Inspired by Alexander Tedeschi’s Master’s thesis “A geospatial analysis of bike share redistribution in New York City”, we decided to explore Citi Bike’s system data and json feed for the year 2015 in order to visualize and analyse where and when bicycles were transferred and bike availability patterns.

We hope that this data exploration tool will serve as an engaging resource for bikeshare users, urbanists, open data enthusiasts, and those who work in public transportation.

Read more details at Urbica's blog post

Back to the project

Bike share trip data: System data, Openbus json data
Map engine — Mapbox GL JS, graphs — D3.js
Routing engine — OSRM, map data — OpenSteetMap
Design & development: Urbica Design, data processing and analysis: Alexander Tedeschi

Route types:
— Outgoing trip routes
— Incoming trip routes
— Rebalancing (x10)
* Only frequent trips and rebalancing routes are shown
Percent of bikes available at stations:
<10%, empty
10 — 30%
30 — 60%
60 — 90% 
>90%, full
Station types (clusters):
Low availability
High availability in afternoon
High availability in morning and evening
Number of docks:
Percent of bikes available  ()
Trips total (outgoing/incoming):
Incoming balancing total:
Outgoing trips per hour ()
Rebalancing per hour ()