There are two great tastes which need to go together: interactive online maps and online trip-planners for public transportation. The former are user-friendly and graphical, but only relevant for driving (or sometimes bicycling); the latter let you plan transit, but are user-unfriendly, because the user cannot see where any of the pick-up and drop-off points are. (This is especially important when making transfers.)
Since most public transportation systems have a steep initial learning curve that keeps away casual or infrequent users, an interface improvement such as this could increase transit ridership by millions of trips per year. But how to do this on the meager budgets of most transit agencies?
People like Adrian Holovaty and MojoDNA Research have made hacks to Google Maps in their spare time, so that the local bus-system map overlays the city street map. This is impressive and admirable, but it isn't what the user ultimately needs. What people need is to be able to type in their start address & end address and have the website show them a map with directions, just like driving directions except for transit, with the route on the map showing where to pick up, where to get off, and where to walk between stations or destinations.
One American city's transit company has almost done this: Portland's TriMet will show you a map of your start or end address with the correct bus stop highlighted, but it won't show you the whole route. Boston-area e-transit also makes an attempt, but falls shorter. (There may be successful European or Asian agencies, please comment if you know them.) Also, TriMet and e-transit use proprietary systems (and old ones at that)--this means they are no match for the modern apps like Google Maps, they're expensive to update and maintain, and their success does not help other cities.
So what the world needs is an open standard XML for getting directions in and out of transit-planning systems.
It can be very simple, just something to read addresses and times and tag them as get on / get off, tags to color-code segments of your trip as being bus / walking / subway / etc, and perhaps tags to include multiple alternative routes for the same trip, or tags to pass messages back and forth when the transit-planner asks you to clarify an address. This way you type your start and end addresses into Google Maps, it throws the data to the appropriate region's transit system, the transit system figures the route(s) and throws some XML back to Google, which then draws you a map with directions (and maybe includes links for alternate routes, which are served up from the same XML file). Then after an initial slightly-painful period of rewriting the trip planners to read/write this XML, and a painless little mapping-software extension like the ones listed above, all transit agencies can have fantastic user-interfaces with no effort (or future maintenance costs associated with interface).
And, as a helpful consequence of universal open standards, different transit systems (say, local bus systems and city-to-city trains) could intercommunicate as well. This would solve another large user-interface hassle with transit.
In Britain there's Transport Direct which provides multimodal route planning and can show public transport stops on a map.
I have often started work on such a system, though I have so far been stopped at the data-acquisition phase, as every time I've tried it in the past, GSM was not available. Now, with GSM, it should be easy. I think this is one of the most obvious uses of open source and should have been done a long time ago. I wonder, how come there is not yet funding for this obvious move? I also wonder, how can society ask for any more open source from people like me pro bono, when our basic needs are not at all being addressed? http://hcvaction.org/ It seems like this governance system is broken. But I support the idea that it would be good to somehow support open source as a society; I think the issue is not technical but social.
www.hopstop.com in NYC works well for busses and subways, though it's not open source, of course.
Actually, it's not so easy to use the major commercial mapping sites for bicycling, either. They don't include dedicated bike paths, and don't let you exclude that great amusement park of the automobile, the Interstate system, which is off-limits to bicycles (and generally interferes with every American's Constitutionally-protected Right-of-way).
How about getting greens, urban dwellers, and pub trans advocates to lean on some of these mapping companies? I think its inexcusable that Yahoo Maps, for one, does not have info on subway stops or buslines.
I tried getting my local transit authority to give me their schedules in csv format- but they refuse to give it out as anything but pdf. They won't even give their drivers any information not pertaining to their routes!
They said they didn't want to be stuck maintaining someone else's application, and would rather buy the commercial package to do it themselves. Even though I agreed to release it as GPL.
My city of Halifax, NS, also does not want to give away any GIS info, so I couldn't get so much as the location of the stops, or the shapefiles for the electoral district.
For all the good Google Map does, it's just as useless as the other companies with regards to bike routes. I don't think we'll have real citizen-friendly mapping for our spaces until our data is in the public domain and the source and services can be modified by anyone with an interest and the skills.
I started writing algorithms for dealing with bus route info and planning efficient trips. If someone wants to deal with GoogleMap or other front-end integration, please get in touch with me- it'll give me motivation to get it polished and finished. If you have csv data of any transit system, all the better as I don't care to do this for my city anymore.
Yes, Open Standard XML and Open *Content*, with Open Source software to drive it. Toss in advancement to the Wikipedia with some GPS, and you have a living map, along the lines of what h2g2 was planned to be.
The Washington Metropolitan Area Transit Authority (wmata.gov) has this.
Here is what should happen: You need to go across town to visit somebody. You enter the destination address into the computer. It proposes a trip that requires 1.5 hours with 2 transfers. You would just give up and go back to your car if you haven't notice the alternative suggestions. One only 50 minutes by using taxi at the last leg. Well worth the extra 5 bucks you think.
Just like most mornings you are getting out a little bit late. You manage to rush to the bus stop just before the scheduled arrival. The bus is nowhere in sight nor do you find any other passenger there. A big anxiety is looming for everything might go wrong. Fortunately a message appears on you GPS enabled phone that the bus is going to arrive 10 minutes late due to traffic. You relaxed a bit. Then the phone tips you off that there is a local's favourite coffee shop just around the corner. Sure, you really hate to miss your morning cup. Finally you are on board with you hot coffee. You got an update that the bus is expected to arrive to the destination 20 mintes late this morning due to heavy traffic. When you talk to you friend last night he think it is foolish to take taxi and offered to pick you up with from the station instead. Good that there is still time to let him know you will be late. The bus is nice enough that you decide to take a nap. Although you have never been to that part of town, you can trust the onboard tracking system to wake you up 1 minutes before arriving to the destination.
Thanks for the links, folks!
wmata.gov is just another text-based trip planner, so it doesn't cut it. TransportDirect.info is impressive because it's city-to-city planning that includes local bus directions, and it has the nicest map graphics, but if you actually look at the maps it gives you to walk from a bus stop to your destination, you'll see that it requires super-powers of walking through buildings. It just draws a straight line. (D'oh!)
Hopstop.com, however, is solidly good. It's even better than TriMet's system. ...Even it could be better, though, by integrating with Google Maps where you can see your whole route (something TransportDirect does), and you get a better interface for panning around, large map, quality of graphics, etc. And as you said, it's not open-source, so it doesn't help any other cities.
London Transport has this-> http://journeyplanner.tfl.gov.uk/user/XSLT_TRIP_REQUEST2?language=en&ptOptionsActive=1
I just tried it out with street addresses I remember and it seems to work. Any regular readers living for a more nuanced analysis?
Sorry, but the maps in London Transport's JourneyPlanner are a hack. They're PDF files, in addition to being separate maps for the start and end (a problem that several of these sites have). They've tried to hotwire the PDF's to simulate scrolling and zooming, but it fails. It's the wrong technology.
I appreciate the further sites posted, but it's proving my point that nothing presently does the job. Hopstop was good; if anything better than that exists, someone please post it.
I did a smoke test on hopstop, and here were my results on my first try:
E 68th St and Park Ave, Manhattan
1st Ave and E 10th St, Manhattan
subway alone: 38 mins.
subway + bus: 42 mins.
Not exactly intuitive results. When I widened my search criteria I was in effect trying to convey that I didn't care how I got there, just as long as it was better- not that it had to include both modes of transport.
Hopstop does a lot of things right, and I'm otherwise surprised at their features. I really do appreciate walking directions.
So the playing field is not very competitive, and a free / open-source implementation could/should be very attractive to some municipalities.
O'Reilly is putting out a book on OSS GIS this August, and my favorite bookstore is to call me when it's out. I guess I'll have a very concrete objective then :)
Jeremy & Others:
I'ld like to share with you current developments in this area: transit standards for data sharing, limitations of existing trip planning software, multi-modal trip planning apps and RFPs out for them, potential to tie into KeyHole (soon to be Google Earth)...
Open Source is definitely the way we want to go.
Please call me.
GIS Manager, TriMet
While not directly on topic, this has brought up an issue that has pissed me off for some time. Most governments have map data directly avail due to city planning, evaluations, transit.
I would also wager to assume that most countries have a 'freedom of information' policy that allows citizens of that particular country the access to government documents.
I haven't verified any other country other than Australia, but while they do allow access, it's with a fee that's unfounded. (100k rural city costing over 1300 AUD alone). There's also stipulations against repurposing the data for your own means. So much for 'ownership via tax'.
I have no issues with companies providing value add to maps and selling that information. I also don't mind companies that self compile that very same data and repurpose it for their own use. I do see it as a 'issue' when data -that is already in digital form (no retrieval costs involved) that there is a barrier to entry for non for profit use.
If anyone knows where someone can get access to free (for repurposing use) map data on Australia, feel free to let me know.
-- I know that 1300AUD isn't much. Though, consider expanding it to cover a state, or a major city like Melbourne or Sydney and you'll see the impact it may have.
So consider yourself lucky that sites like these thrive where you are, some can't even get started in the first place.
Check out http://www.htm.nl/smartsite.dws?goto=23 , a good start
In the NY/Tri-state area, there's a site called Trips123 that consolidates almost all of the subway/bus/rail traffic throughout NY, NJ, and CT. There's walking directions to and from stops and even transit schedules. I find it very useful. www.trips123.com
Look, all it takes is a few hobbyists with GPS's to do the exact same thing to create the same information that exists behind locked doors. That's how GNU/Linux was written.
They can copyright their works, but they cannot copyright the planet. Think about it. And using a Creative Commons license, you can use a non-commercial license so they cannot use your work - without an agreement with you.
In Finland we have this thing called www.matka.fi .