I originally planned to do a talk on this, but sadly it didn't get picked up. Because of that, I haven't really dug into the data lately, but I'm planning to give it a proper go in 2026.
It’s roughly 100 GB of raw compressed data—the GPS pings for every available Nextbike every 3 minutes.
By polling regularly like this, you can extract trips; I’ve counted 19 million up to October. By using an OpenRouteService server, you can guess which routes people are taking. It’s not a perfect methodology, but it's really interesting for urban planning insights.
Next steps could be correlating it with weather data and so on. If you have ideas, let me know! I’d also be happy to find a co-speaker for the eventual talk.
Also, if you have tips on how to handle this data responsibly/ethically, I’d love to hear them.
Context: Here is an analysis I did in 2022: https://pzauner.de/?link=%2FNextbikefahrten%20in%20Karlsruhe
contact info: Mail: 39c3@pzauner.de signal: pzauner.37 Matrix: @pzauner:matrix.org