Thibaultmol 🌈

If you're ever taking trains in Europe, please checkout Chuuchuu:
https://www.chuuchuu.com

It's a great site that uses historical train delay data to better estimate good train rates and realistic transfer windows between trains and such.

It seems like a must have to me but doesn't seem to be that popular yet

EDIT: they're on fedi now! Go follow: @chuuchuu

chuuchuu - find reliable train journeys

Find train journeys with the lowest chance of delays, cancellations and missed transfers.

Mar 3 at 8:12pmWeb
Show threadThe DoctorMar 4
@thibaultmol
Show threadBeady Belle FanchannelMar 5
@drwho @thibaultmol only acceptable answer cc @RosaCtrl
Show threadMer-fOKxTOwlMar 4
@thibaultmol on my most taken route >10% of all trains get canceled :/
and it does not even have a change, and only takes ~1h10min
Show threadjoëlMar 4
@thibaultmol this looks very cool, but the 100% score on Luxembourg-Brussels is … concerning.
Show threadNerdfallmanagementMar 4
@thibaultmol @evawolfangel wäre das nicht was für dich ;-)?
Show threadRaphael WimmerMar 4

@nerdfall @thibaultmol @evawolfangel

Bitte nicht, sonst haben bald auch alle jetzt noch pünktlichen Züge Verspätung.

Show threadEva WolfangelMar 5
@RaphaelWimmer
😜
@nerdfall @thibaultmol
Show threadRaphael WimmerMar 4

@thibaultmol

Cool!
For Germany, Austria, BeNeLux, Italy, Denmark, and Slovenia, there is also https://www.zugfinder.net/en/start

(some features require a paid 'Pro' account - price is fair IMHO)

It's all about trains. Current train positions in Germany, Austria, BeNeLux, Italy, Denmark and Slovenia

Zugfinder displays the current position of trains and offers extensive delay statistics.

Show threadThibaultmol 🌈Mar 4
@RaphaelWimmer i don't think it has the same route planning functionality based on those delays like this tool does though. I might be wrong
Show threadwortwartMar 4
@thibaultmol Nice -- but I tested it with my next major trip (Berlin - Palermo) and after some thinking all it gave me was an error message.
Show thread~RiWe Mar 5

@thibaultmol
"Find reliable trains"
Germany, no results...

#deutschebahn

Show threadPeter Juelsgaard GjaldbækMar 5
@thibaultmol it doesn't quite seem developed enough yet, at least for my use of finding trains from Denmark down through Europe (the search comes up empty with an error message). However, I will definitely keep an eye on it, as the transfer timings are one of my main gripes, when using other planners
Show threaddusoftMar 5
@thibaultmol Very limited data for now, maybe third of Europe.
Show threadAurora 🏳️‍⚧️Mar 5
@thibaultmol me, regular person browsing just completely ordinary fedi timeline, reading "if you're trans in Europe, please checkout chuchu." 😅
Show threadCyberGwenMar 5
@thibaultmol Not really required in Switzerland though. Average delay time is pretty much zero here 😂
Show threadLeonardo Ferreira FontenelleMar 5
@thibaultmol @anna no plans of using the service, but the name… 💯
Show threadchuuchuuMar 12
@thibaultmol Thanks Thibault for the mention, it generated a bunch of traffic! Can't wait to build out chuuchuu into the best train travel assistant!
Show threadSMillerNLMay 24
@thibaultmol @chuuchuu I’m checking this out even if it’s just for the name.
Show threadBig GeorgeMay 25
@thibaultmol @chuuchuu great name....
Show threadNewbyteMay 25
@thibaultmol @chuuchuu 48% chance that I'll miss my transfer on my upcoming journey, great!
Show threadchuuchuuMay 25
@newbyte Good luck! 😅
Show threadOskar im KellerMay 25

@thibaultmol @chuuchuu

There is next to no information on the website about who is behind the project. For me that is a bit of a red flag. Also it looks like it is more geared to casual #train customers.

Travelling mostly on #DeutscheBahn I'm used to calibrating my trip with sites like https://bahn.expert which also offers historical data of trains and their delays.

Bahn Experte

Dein Begleiter um Stressfrei Bahn zu fahren. Sucht die besten Informationen aus allen Quellen um dich ans Ziel zu bringen.

Show threadchuuchuuMay 25
@OskarImKeller Hi Oskar, so far I've just built this as a trial project. At the moment we're focusing on our data collection and then we would continue building out the app. Let me know if you want to know more about us!
Arne
Show threadThibaultmol 🌈May 25
@OskarImKeller @chuuchuu
- it uses the same api as bahn.expert probably
- It's not really mentioned anywhere but this is made by Arne Nys (a mobility design consultant who's made various things). It's not anything sketchy, but I agree that it would be good to mention his name somewhere
Show threadOskar im KellerMay 25

@thibaultmol @chuuchuu

Thanks for the explanation.

Show threadLucMay 25

@thibaultmol tried a domestic route to Cologne and a domestic route to Eindhoven. Chuuchuu scores are 79 and 99, respectively. That matches my experience, so n=2 I declare this reliable :D

(Or, at least, better than not having any idea when you visit a new place)

Will bookmark this! On F-droid I saw Bahn-Vorhersage before, but that's only for Germany and often doesn't have data. These two should probably join forces!

Show threadOliver Blanthorn6d ago

@chuuchuu this is a very cool idea but i think it needs a little more work on the scoring

it would be absolutely fantastic if this journey really worked but in reality you have an extremely high chance of missing the connection, not to mention that at some stations the SNCF refuses boarding to anyone arriving five minutes or less before departure

and if you miss it the next train is at 1pm : ) 🇫🇷🇫🇷🇫🇷

@cycling_on_rails

Show threadCycling on Rails6d ago

@bovine3dom I'm surprised this connection didn't get a percentage of missed transfers, others have it. Here it would be a 15% miss chance. Also, SNCF only measures delays in 5-minute increments, so extra padding should be added accordingly.

I'm also not convinced by a 14d rolling average to measure rare events. Looking back more than a month and/or adjusting for seasonality would be more accurate.

But despite these remarks, it'd make a cool product! @chuuchuu

Show threadOliver Blanthorn6d ago
@cycling_on_rails i suspect the two-week window is for practicality - you only have to scrape two weeks of data rather than ~4 years if you want seasonal adjustments
@chuuchuu