Morning Session: Transit

1 view
Skip to first unread message


Feb 20, 2011, 12:16:17 AM2/20/11
to DataCampSEA
Brian Ferris, getting PhD in Computer Science, working on One Bus Away
Dave Albreight, Wishery, eSwing Golf, Pure Home
Mukund, software engineer at Facebook. Converting unstructured data
to structured data.
John W,
Andrew Morton, UW Informatics grad, working with Drupal. Library
computer project.
Haley, Transportation Management for City of Bellevue.
Tim Bond, UW Informatics grad, transit enthusiast.

Human guides for how to use transit. Advice given to someone would
differ from advice given to someone Seattle, especially for things
like pass sales.

What are problems we’d like to see fixed, and how can we use data to
solve those problems?

Biking in Bellevue is a challenge since Bellevue was designed for

There are some static maps, and Google has interactive maps. But how
do you improve this data? Talk of a “ride share” for groups of
bicyclists to help newbies—have a newbie commute with a group of
experienced bicycle commuters. There are apps (mobile and otherwise)
that can help get this information out, but finding them is a
problem. For instance, how would one find a map that say a Bellevue
resident designed? There are multiple ways to get from point A to
point B, how do you find the best way?

Upload your GPS trails and/or browse others’ at Trails
broken down by biking, walking, driving, etc.

If you are trying to get from point A to point B and it seems to take
a long time via bus, how do we help the county/city plan new routes?
There are social equity issues with this. We’re always going to want
more service than we can afford. 40/40/20 rule. A Transit Master
Plan was recently published. Help people understand how transit
policies are made. There is a large segment of the population that
don’t have smartphones, so how do you help those people? Agencies
can’t always promote one project as that shows favoritism.

CityGoRound has a visualization of what cities have open data and
which don’t. NYC is a newcomer. Brian is working on a transit trip
planner that uses real time data.

Getting access to open data: agencies should release some of this
data, such as passenger count. Another solution is to PDR (Public
Disclosure Request, sometimes known as FOIA, Freedom of Information
Act). Sound Transit (
Us.xml#records) and Pierce Transit (
pdf_index.htm) have forms on their websites; for King County Metro
contact Craig McMurdo (
EmployeeDetail.asp?EmpID=37709). Sometimes data from the agencies is
inaccurate or inconsistent, so how do you get this data fixed?
Government agencies are overworked and time crunched. Much discussion
about route usage data and opening up this data for all to see/use/

How do you crowd source this data? Examples: passenger load, bike
rack usage. “Rate my route”. For instance, commuting from Ballard to
UW: 46 provides a more enjoyable ride than the 44. Google Maps
calculates driving cost VS transit cost, but how do you include real-
user reviews into that?

What-if analysis? i.e. avoid peak fare, shortest trip, etc.

One Bus Away’s best kept secret: Explore tool. Put in your starting
address and a category, i.e. “taco stand” and a time and it’ll show
you the best places and the bus routes that will take you there. Also
includes time constraints, such as “taco stands I can get to from the
UW in 30 minutes or less by bus”. Walkscore (http:// can help finding a place to live.

One problem is when external things go wrong: snow, bus breaks down,
subway isn’t coming, etc.

Go 520 project. Real-time ridesharing project. For help with
upcoming tolling. Powered by Avego.
(editor’s note: it’s worthless unless you buy into the Apple kool-aid)

Why do we have a peak hour fare? Dynamic pricing. Farebox revenue
does not pay for 100% of the operating costs. Service may cost more
to provide in peak periods, some may have to do with deadheading.
Other reason may be because they “can”. Sometimes things are priced
high to encourage those that are flexible to not use it. Tolls on 520
are a great example.

Transit data is usually exchanged in GTFS (General Transit Feed
Spec). Some places to get GFS: GTFS Exchange (http://www.gtfs-data- There isn’t really a standard for real-time data, but
SIRI is becoming popular. See also City Go Round (http://

Another idea: using past data to predict the behavior of future
trips. Such as trips on Friday during peak may take longer than the
same trip on a Monday during the same time.

King County Metro is upgrading their signpost-based AVL system to GPS,
scheduled to complete by April 2012. Along with this will come
automated next-stop announcements and displays.
Reply all
Reply to author
0 new messages