Transit Data, Open Government, and Vision

Today, we have a guest post from Bruce Roberts on a recent transit data workshop hosted by King County Metro.

Last Thursday evening (October 22nd), King County Metro (KCM) held a workshop about its transit data APIs aimed at the technical community. Approximately 50 people attended, with about 30 members of the technical community and the rest primarily technical staff from King County Metro, Sound Transit, and Portland TriMet. The workshop consisted of a brief introduction, a panel discussion, a breakout session for brainstorming, and a summary report of the brainstorming sessions. There was additionally ample time during breaks for discussion with individual representatives from the government agencies.

KCM currently provides schedule, route, and real-time bus location information. The real-time data on its buses is done through an automatic vehicle location (AVL) signal-based system. The buses essentially signal where they are in a route based on how far they have come. This doesn't work well if the bus is rerouted for construction, or the route changes due to emergency conditions. In addition, approximately ten percent of KCM's buses are equipped with automated passenger counting systems, and these buses are currently rotated through different routes to collect sampled passenger data. I don't think this passenger data is easily available to the developer community yet.

Two changes to the KCM system are going to be rolled out soon - KCM is planning on deploying a GPS-based system in their buses next year which will allow for tracking actual locations of buses, rather than inferred locations. They also plan to begin providing the schedule and route data using an emerging standard based on the Google Transit Feed Specification. It wasn't clear to me whether there is an emerging data standard for vehicle location data, or when passenger count data might be made openly available.

KC Metro is clearly very open to working with the developer community. The staffers repeatedly represented that they wanted honest feedback about their efforts to date, and guidance for short-term and long-term updates to the program. One staff member said "it's the public's data." I found this pretty refreshing from a government agency. It was very clear that they wanted an open exchange, and I believe they got a lot of very useful feedback from the community. The technical community had a lot of feedback - this fell into a number of general areas:

1. Continue to support the development community, though further workshops, online tools, whatever they can. If this means finding an existing transit data forum or group that can be used for the Puget Sound community, that's fine, particularly if that community can be set up faster than one supported directly through KCM resources.

2. Focus on getting the community as much data as possible, including historical data. Don't assume that only schedule and real-time data are useful - let developers, entrepreneurs, and researches figure out ways to use whatever raw data you have. The technical community is happy to drink from the firehose.

3. Spend time making sure the data you do provide are correct and the raw data APIs are well thought out.

4. Don't spend time worrying about supporting the old legacy AVL APIs - move as quickly as possible to the GPS-based data. This one seemed to come as a pleasant surprise to some of the technical staffers, who I think are used to working in a different environment with different constraints than much of the technical community in attendance.

5. Work to simplify the legal requirements for using this data. Portland TriMet has had great success with this, but admitted that it was a learning curve for their lawyers to accept providing data with limited (if any) restrictions on its use. KCM will likely need to go through this process as well.

6. Work to get us data from other systems as well - ferries, light-rail, whatever is available.

One of the developers at the workshop painted a great vision, and I think it was a vision generally shared by all of the participating parties at the meeting. He wanted to be able to step off an airplane anywhere in the world, access transit information from his mobile device for all the appropriate transportation options in that city, and use that information to get him to his ultimate destination. I believe this is a wonderful long-term vision, and I believe that we'll continue to make significant progress towards that goal. It will take time and effort (but what worthwhile thing doesn't?). This workshop demonstrated that there is interest both within government and and within the technical community to work towards that vision.