PT Database and GTFS Feed for Metro Manila, Philippines

[tay...@itpworld.net]

Hi All

I think I may have met some of you at Open Plans’ Transportation Camp
in Washington back in January, and I may have mentioned a World Bank
funded project to create a public transport database and GTFS feed for
Metro Manila. Since then we have been working on behalf of the
Philippines Department of Transport and Communications, and a number
of public transport agencies based in Metro Manila, to collate all of
the available public transport route and schedule data.

The project’s aim is to build a simple (for a range of reasons, likely
to be MS Excel-based) data collection tool which will be deployed in
Metro Manila as a means of collating public transport data so that it
can be consumed by a simple open source GTFS-creator software
application, which is being developed through the project. Our client
at the World Bank has funded this project because they believe a suite
of simple data gathering and GTFS creation tools can greatly benefit
other developing cities with largely informal transit networks as they
work to rationalise and regulate them.

The first interesting problem we have encountered is that a lot of the
data simply doesn’t exist, and so we will be working with the relevant
agencies to build capacity and collect public transport route shape
information that it can be maintained in GIS and form the basis of a
GTFS feed. The data collection activity to fill the various agencies’
knowledge gaps about the routes they franchise is being planned for
June and July of this year.

The next challenge relates to the informal nature of public
transportation in Metro Manila. The situation here is similar to
other cities in which we have worked on behalf of the World Bank in
that there are:
• Informal stopping points – for buses and jeepneys the whole route is
theoretically a stop, as the driver’s profit-maximisation incentive
means they respond to flag downs at any opportunity along their
designated route.
• No transit schedules – Drivers of buses and jeepneys are not
required to operate to a schedule, with most service provision focused
around the morning and evening peak periods because this is when there
are the most fares.

While we can fool GTFS into believing there are schedules and stops,
based on average headways and commonly-used formal and informal
stopping points, I wondered whether you are aware of any cities where
GTFS has been deployed in a similar context?

The final challenge we face relates to the use of distance-based
charging. In Metro Manila, and across the Philippines, fares are
typically a fixed rate for the first 4km of a trip, and then increase
per km travelled thereafter. Having extensively reviewed the GTFS
spec we believe it should be possible to do include this fare model
using the shapes files to calculate distance-travelled along routes,
but wondered whether you know of any code that has been/could be
developed to make it work?

I warmly welcome any ideas or contributions you can provide, and would
appreciate the opportunity to discuss this project with you in more
detail.
Best regards
Neil

Neil Taylor
Principal Consultant
Integrated Transport Planning Ltd.
32a Stoney Street
The Lace Market
Nottingham
NG1 1LL

T: +44 (0) 115 988 6903
UK Mobile: +44 (0) 7780 710152 (UK)
PH Mobile: +63 (0) 947 4856017 (Philippines)
E: tay...@itpworld.net
W: www.itpworld.net
Skype: neil.taylor42

[David Turner]

I think we did meet at TransportationCamp. It's good to hear from you
on the list.

I've never heard of a tool that supports any of these three features.
However, the project I work on, OpenTripPlanner, supports pluggable fare
models, which would easily handle distance-based fares. I expect I
could have a plugin written in under an hour.

We're also working on predictive availability of bike shares; extending
that to buses and jeepneys would not be impossible. Is there historical
or realtime data available on the bus/jeepney locations or frequencies?

I think it would be possible to add support for arbitrary stopping
locations as well. But it would not necessarily be straightforward.
I've written code that snaps hand-drawn or GPS-generated route
alignments to the street network, which could be a good first step.
Using this mapping, we could allow the transit network to piggyback on
the street network.

If you're interested, come talk to us on the opentripplanner-developers
mailing list and show us the data that you have. Or contact
kw...@openplans.org if you're interested in a more formal arrangement.

[tay...@itpworld.net]

Thanks David

I have replied to you offline following up on your message.

Cheers
Neil