/obike

Working with the (strangely) very open oBike API

Primary LanguageJupyter NotebookMIT LicenseMIT

Understanding oBike and it's business case

Working with the (strangely open) oBike API

TESTETSTETTSTETST Juli 2017, the bike rental service oBike from Singapore started in Zurich. It lets users rent whichever bike they feel like, simply by scanning a QR code. This frees users of the hassle of bringing bike rentals back. The service is sensationally cheap, charging only 1.50 Swiss Francs per half an hour. The bikes themselves are therefore understandbly cheaply produced, including the actual technology involved. The API e.g. showing the real time bike location is openly available. The following analysis examines the data published in oBike API (which has since been modified by the company).

The piece appeared in the "Tages-Anzeiger" on September 2017. It showed how much the companies currently earns, how long people spend riding the bikes around the city and where and when they are doing so.

These were the steps involved:

  1. Pulling down a snapshot of the data every 1/2 hour. Here's the raw data.
  2. Analysing the files and creating a origin and destination lat/longitude.
  3. Sending the data to the Google Distance Matrix API.
  4. Working with the final results from the Google Distance Matrix.