Try it out:
http://www.bizave.com/cgi-bin/busrb.cgi
You can add your own readerboards from there, with up to eight bus
lines on one board. Pick any stops you want. The readerboards are
saved on your computer alone, though you can share readerboards with
others.
For example, here's one I made for the #14 and #15 in SE Portland, at
39th Ave, heading to Portland:
I sent a note to TriMet - they have been playing with this.
Andrew
--
----> Portland, Oregon, USA <----
*******************************************************************
----> http://www.bizave.com/photos/ <---- Andrew's Photography
----> http://www.moviepundit.com/ <---- Andrew's Movie Website
*******************************************************************
To respond by EMAIL, please DO NOT hit "reply". Email address
scrambled for security. Instead, please visit the web page:
Andrew> TriMet recently started making real-time bus arrival times
Andrew> available on their web site. This is terrific, but I really
Andrew> wanted something that would show more than one bus, like the
Andrew> readerboards they have for a few bus lines and for the MAX.
Andrew> So I wrote a cgi web page to do so, using TriMet's web data.
This is very cool and useful for transit users, but one step better
would be to be able to pick a destination and use the real-time (and
historical) data to estimate the fastest choice to a particular
destination. For example, there are two bus lines that go near my
home, but that take different routes downtown. Which one will get me
there first?
--
Russell Senior ``I've seen every kind of critter God ever made,
sen...@aracnet.com and I ain't never seen a meaner, lower, more
stinkin' yellow hypocrite than you!''
-- Burl Ives as Rufus Hennessy
: Andrew> TriMet recently started making real-time bus arrival times
: Andrew> available on their web site. This is terrific, but I really
: Andrew> wanted something that would show more than one bus, like the
: Andrew> readerboards they have for a few bus lines and for the MAX.
: Andrew> So I wrote a cgi web page to do so, using TriMet's web data.
: This is very cool and useful for transit users, but one step better
: would be to be able to pick a destination and use the real-time (and
: historical) data to estimate the fastest choice to a particular
: destination. For example, there are two bus lines that go near my
: home, but that take different routes downtown. Which one will get me
: there first?
I don't have enough info from TriMet's web site to create such a
thing, anyway, but I think there are some difficulties with doing
that. You can't predict things like accidents, delays like handling
wheelchairs, etc. So I'm not sure how much I would rely on some
prediction about expected travel time.
What TriMet *ought* to be able to do is publish average point-to-point
times between any given stops, based on past data. Then you can
simply pick which bus you prefer based on it's average travel time and
convenience to your source or destination.
Andrew
--
----> Portland, Oregon, USA <----
*******************************************************************
----> http://www.bizave.com/photos/ <---- Andrew's Photography
----> http://www.moviepundit.com/ <---- Andrew's Movie Website
*******************************************************************
To respond by EMAIL, please DO NOT hit "reply". Email address
scrambled for security. Instead, please visit the web page:
Andrew> This is very cool and useful for transit users, but one step
Andrew> better would be to be able to pick a destination and use the
Andrew> real-time (and historical) data to estimate the fastest
Andrew> choice to a particular destination. For example, there are
Andrew> two bus lines that go near my home, but that take different
Andrew> routes downtown. Which one will get me there first?
Andrew> I don't have enough info from TriMet's web site to create such
Andrew> a thing, anyway, [...]
Of course not, but with the data they obviously have, it could be feasible.
Andrew> but I think there are some difficulties with
Andrew> doing that. You can't predict things like accidents, delays
Andrew> like handling wheelchairs, etc. So I'm not sure how much I
Andrew> would rely on some prediction about expected travel time.
But since it is intended to be an estimate compared against other
estimates (i.e. the various alternatives) computed in precisely the
same manner, then it could still be useful.
You can't predict accidents and other delays in the "when your bus is
coming" either. It isn't a guarantee, it is an estimate.
Andrew> What TriMet *ought* to be able to do is publish average
Andrew> point-to-point times between any given stops, based on past
Andrew> data.
But that isn't enough. You have all sorts of "time of day"
variability and so forth. I'd prefer they publish all the
time/position data and let others build applications from it. Now
_that_ would be _really_ cool. I don't understand why they would
consider the data proprietary. They are a public agency. They aren't
in direct competition with anyone. Whatever makes their system more
useable to riders makes them more successful.
> This is very cool and useful for transit users, but one step better
> would be to be able to pick a destination and use the real-time (and
> historical) data to estimate the fastest choice to a particular
> destination. For example, there are two bus lines that go near my
My impression is that the trip planner that TriMet has on its web site
(getting from A to B) is currently based, at least to some extent, on what
is happening on a particular route and bus location.
--
-Glenn Laubaugh
Personal Web Site: http://users.easystreet.com/glennl