VeryFurryFerryFairy – A Case For Open Transit Data

The ferry terminal at Buiksloterweg with the display showing the time until next departure. The ferry I just missed can be seen speeding towards Central Station in the background.

I take the ferry from the center of Amsterdam to Noord, the part of the city on the far end of the IJ where I live, at lot. For cyclists and pedestrians, it is the only way to get from Noord to the city and vice versa, so they are used a lot. There is a total of five ferry routes, all with different schedules: weekdays, weekends, round-the-clock or restricted. Too much to remember even for the seasoned ferry traveler. Consequently, you frequently find yourself waiting for fifteen minutes or even a half hour, especially late at night when all you want to do is get home and crawl into bed. Wouldn’t it be great if you could just check your phone for the time until the next ferry just before you head out?

Open Data

A ferry departure app is such a simple and practical idea that it makes you wonder why it doesn’t exist yet.

The answer, I believe, is really simple: because there is no easy access to the schedule data for developers. Sure, you can scrape the GVB website or just copy the schedule information into your app or website. But frankly, that’s just a lot of work, and developers are lazy (among other things). When you, GVB Amsterdam or any other transit provider, open up that data, you unleash the creative power of your travelers, your customers. They feel the itches of everyday public transit use. A lot of them have the skills and the ideas to scratch those itches and create useful mobile applications, great websites, cool visualizations or even physical products that help other transit users to make their lives easier. This video tells that story better than I ever could:

The Fairy

The Very Furry Ferry Fairy running on an Apple iPhone

In the mean time, we’re stuck with static information on the transit authorities’ web sites, and a national public transport journey planner web site that really only works well for technology-savvy people with a lot of patience and perseverance, and does not cater at all to specific use cases like my ferry itch.  The imminent addition of some Dutch transit routing information to Google Transit (links to Dutch page) may seem like a step in the right direction, but providing information to Google is not quite equivalent to opening up data: transit authorities share their data with Google exclusively, and once it’s in, it stays in. I am sure a lot of people will find Google Transit useful – I know I will use it – but it does not allow for the creative ecosystem that true open data would enable.

To give you a tiny sampling of a simple but useful service that could be built on open data, I give you The VeryFurryFerryFairy. Open the link in your mobile phone’s browser, select your favorite ferry and bookmark – and you’ll never be waiting around on the dark, cold, rainy ferry dock again.

And Now?

I created the VFFF using static schedule data that I encoded into the web page source. This works quite well because the ferries generally stick to their schedules well, there’s almost no disruptions in the service and the schedules themselves don’t change all that often. For any more complex real world use case, we would definitely need an transit data API that exposes schedule information, but also temporary schedule and route changes and live vehicle positions. The latter are crucial for accurate departure time predictions on crowded urban transit networks that are prone to frequent disruptions. The value of  live data was clearly demonstrated when the London Tube Feed, offering live vehicle positions for the entire London Underground network, had to be suspended due to capacity problems following overwhelming demand.

In the Netherlands, the value of open data is not appreciated yet by the authorities holding that information. There is an  initiative towards a national data infrastructure for public transit data, NDOV, but that won’t yield any usable results until 2015. That is a lifetime in a world where people adapt to new ways of interpreting and consuming information! It is my firm belief that this topic should be addressed bottom-up. Get some locals together, camp-style, to scratch a few simple itches that involve transit data. Involve some local transit authority and local government reps and see what you can build – prototypes, ideas, cooperation – in a day.

So that’s what we’re working towards. See you soon at TransitCamp020.


3 thoughts on “VeryFurryFerryFairy – A Case For Open Transit Data

  1. Being a ‘Noorderling’ myself, I had the same itch, and scratched it. I also hardcoded the timetable into my code, hoping the scheduals won’t change often nor any time soon. I also added (static) info from to account for holidays (sunday schedual).
    Both data sources should obviously really be open data sources.

    My ferry solution ( ask for your location and destination on a google map, than plots routes via all ferries (using google directions), calculates total travel time for each option and returns the fastest ferrie. Its a (mobile) website for now, an android app will hit the market soon.

    For experienced ferry travelers though, I think the VFFF is a simpler and faster solution. Most ferry travelers know their alternatives route lengths, they just need the departure times, fast and easy.

Comments are closed.