special sneak preview of the tool today.
(Photo: Michael Andersen)
This guest post is by Michael Andersen of Portland Afoot, PDX’s 10-minute newsmagazine and wiki for transit commuters.
Already known internationally for its top-notch website and pioneering use of open data, Portland’s transit agency is about to take another leap toward making low-car life easy and intuitive.
With the launch of its new Regional Trip Planner – that’s the link to the latest version, which will go live on TriMet’s website tomorrow – TriMet will be the first American transit agency whose website uses open-source software to plan trips the way low-car humans do: by figuring out the best combination of bike, train, bus, foot and even, eventually (gasp) automobile.
“Multimodal functionality is a highly requested feature that not even Google can offer,” said Bibiana McHugh, the TriMet data specialist who led TriMet’s three-year effort, at a presentation Friday.
Using the interesting new “bicycle triangle” tool developed by TriMet’s team, you’ll be able to choose a balance among speed, safety and elevation changes, letting you customize the trip to your preferences and energy level.
Carolyn Young, TriMet’s outgoing executive director for marketing and communications, predicted that the new planner “is going to be the envy of the transit industry.”
Last week, Jonathan laid out the basic functions of the new planner, which was funded by two grants from Metro and executed by a team of in-house developers and interns and a contract with Open Plans (a nonprofit you might also know as the publisher of Streetsblog). Some new nuggets we learned Friday:
- The trip planner will soon include bike parking: TriMet bike and rides, city bike corrals and even bike staples.
- It already features Zipcar locations right on the map, and might eventually be able to tell you whether a particular Zipcar is available. To find this, click “carshare” in the lower left of the new map.
- Within the next year to 18 months, partnerships with nearby agencies C-Tran, SMART and CHERRIOTS will let you plan transit and bike trips all the way from Battle Ground, Wash., down to Salem.
- In two years or so, you’ll even be able to build an auto trip to the nearest park-and-ride into your trip plan. TriMet is in talks with the Portland Police Bureau about gathering data on street directionality, speed limits and other necessary details.
- Third-party applications – the sort that could, for example, let you use your mobile phone to calculate the flattest bike route over Alameda Ridge – can’t yet plug into the system. That’s one of the many features that McHugh said will be introduced in the coming months, with new features every couple weeks.
The new trip planner won’t immediately replace TriMet’s familiar, proprietary version; TriMet expects to run them both in parallel for “three to six months,” McHugh said.
In the meantime, McHugh’s eager for input from early users. Have at it.
Portland Afoot’s October issue visited Silicon Valley to reveal seven secrets about the future of carsharing. BikePortland readers can subscribe for $10 a year with discount code BIKEPORTLAND.
Thanks for reading.
BikePortland has served this community with independent community journalism since 2005. We rely on subscriptions from readers like you to survive. Your financial support is vital in keeping this valuable resource alive and well.
Please subscribe today to strengthen and expand our work.
How is “safest” determined?
I had it suggest that I ride down Division, if I selected fastest, but even 1% safest would make it switch to a bus on division. I guess it considers the bus faster than taking a bike boulevard, but you can’t compare routes easily, like google, and it does not list wait times like the current tracker.
Good question. “Safest” is based entirely on the size of the road, as recorded in Open Street Maps. It’s not crash rates, traffic counts, or anything else.
There’s a bit more about this on our wiki: http://portlandafoot.org/w/TriMet_trip_planner
Real-time arrivals and drag-to-compare routefinding are both on the feature list for the next few months. Not sure about wait times specifically.
I apologize, “size of the road” was an oversimplification. The road category includes such things as bike facilities and its role in the transpo network. See here for a detailed explanation from one of the tool’s creators:
http://portlandafoot.org/w/TriMet_trip_planner#How_.22safety.22_is_determined
“Safest” appears to be determined by the RLIS:*=* tags that the TriMet team provided, which are in turn based on Metro RLIS data and Bike There! data.
It’s true that in certain cases RLIS:bicycle=* tags have an influence on the “Safety” routing, but as Mele explains in the Portland Afoot Wiki the majority of the weights that are applied under this setting are based on the presence of bicycle infrastructure (tagged as cycleway*=* in OpenStreetMap) and street type (tagged as highway=*).
I can’t get this to work. I enter all my info, but there is no “engage” button anywhere on the page that I can discern. How do I get it load the data so it will actually show me my trip options?
Ahh…found it. I am using my net book; some websites can’t handle the condensed aspect ration, so I had to zoom out to reveal the button. FYI…
I just tried to plan a route from my home near Alpenrose dairy to the Beaverton Library. Search form worked fine, then got a “Loading” icon for about 13 minutes (literally) before I bailed on it. Other content is loading fine.
OK. It worked on the second try. Nice tool. Unfortunately, it still tells me I have to walk a mile (literally) to catch a bus, and that my mid-length trip will take an hour one-way. (Walk from Maplewood to Garden Home, Bus from Garden Home to Washington Square, Transfer to bus from Washington Square to Beaverton). I’ll try a route to Hillsdale. Maybe Trimet is telling me I’m using the wrong library.
What route it takes also depends on the time of day you’re asking it to route for, if transit it involved at all.
Planning trips by bicycle or “bike and transit” not possible from my home (SW Portland)to either Beaverton or Hillsdale. Error says maybe “outside of map data area”. I guess I will say that I hope it works well for the majority of users, as it seems like a neat-o tool that doesn’t really work for me.
Hi Machu (and anyone else having trouble with the new trip planner),
Thank you for trying our new trip planner. The Portland Regional Trip Planner is in beta release and may still have bugs, but we hope to get feedback from customers so we can improve its performance. We apologize for any problems you encountered. Please send us an email at rtpfeedback@TriMet.org with your trip information, so we can figure out what the problem is and hopefully get it fixed!
Thanks,
Mele Sax-Barnett
TriMet RTP Intern
Cool, ah, it mostly works, that is to say it found some places and not others, or it provided routes that I didn’t consider fastest of safest. But it’s a great start or even mid-point. I’m certain it’s hard to build this stuff, so I don’t know what the challenges are on the back-end. The UI looks good. I tried to route from my home on N Williams to Fed-X, and it only provided me one option, which was on Swan Island (where I want to go), but the destination was not the exact location of the Fed-X drop-off spot. Like Google Maps it provides routes, but if one is not familiar with an area, then the route finder does not always locate the best routes (fastest/safest). Better to have something you can use to get 80% closer to knowing your route and then tweak as you experience the ride, then nothing at all. Or maybe it better to go blind and have to way-find by landmarks and odd features, always leaving and hour early to allow for wrong turns.
Worked great for me. There are tabs at the top that you can use to plan trips by bike only, bus only, and bike-and-bus and then compare routes. The elevation chart is helpful, too – Google doesn’t have that.
It’s important to note that this is a public beta still in development, and not the final product, so don’t judge too harshly, and definitely email trimet at rtpfeedback@TriMet.org so Bibana and the gang can get an idea of what hangups people are running into.
Why did they spend all this money for new software when there was nothing wrong with the old software? Is this a case of “spend it or lose it?”
Jim, I’m not here to defend TriMet’s spending priorities, but this will replace a licensed software package for which TriMet has been spending many thousands of dollars (I’m not sure of the figure) annually. The new product also offers features the old does not, such as bike trip planning.
Now that we’re discussing costs, does anyone *know* what the total cost to build this new trip planner has been and/or will be? Is there any sort of current accounting or a projection anywhere (can we FOIA this info)?
I mentioned in another comment that open source is not the same thing as free. When *all* costs are considered, the open source option could even be more expensive.
And whether this open source option is better in other ways is certainly debatable. For example, has anyone considered why it is that Google doesn’t offer this type of multi-modal trip planning? It may not be all that useful in a practical sense. Or, it may be useful to certain people in certain situations but not worth the cost overall.
The Business Software Alliance called. They want to know where you want your shipment of astroturf delivered.
So, because I ask a question you don’t like, I’m an astroturfer? I’m sure that attitude’s useful in helping get everyone on board.
Anyway, you didn’t address any of the substance of my comment. Would you care to do so?
Good questions, Peter. The main source of funding was a one-time $75,000 grant from Metro‘s Regional Travel Options grant program, which comes ultimately from a package of federal money designed to help metro areas meet requirements of the Clean Air Act by reducing drive-alone traffic.
McHugh said a second Metro grant helped pay for this, too. I’ll ask her its size and post to our wiki’s section about the grant’s funding.
For comparison, McHugh said licensing rights to the legacy street map alone (not including the trip-planning software) cost TriMet $25,000 annually. That was the cost, she said, of hiring four interns to permanently improve Open Street Map by adding Metro data.
That said, as a website manager myself, I definitely think you’re right that open-source software carries its own costs and is not always the right or cheap solution.
You do know it’s an open source project, right? The savings in licensing fees alone mitigates any manpower costs for the project.
And the new tool is much more useful.
This is just patently false. That’s not how open source works. Open source does not always equate to a lower total cost of ownership when all costs are considered. I would boldly assert that anyone who argues otherwise (disregarding anecdotes and special cases) doesn’t know what they’re talking about.
I’m so glad the “Longest walk” parameter goes out beyond one mile. That was always a headache for me when trying to combine trimet/bike travel with the old trip planner. I’m more than happy to bike four or five miles if it will save me a connection or a long wait.
This tool seems incredibly useful. Great for planning longer trips than in my neck of the woods, should help me avoid nasty hills! And its combining transit and biking… finally!
You missed the point. Where were you when a commercial package that can’t be fixed and that doesn’t do everything TriMet’s customers need it to do was selected? Why is changing to an open platform that does what it needs to and can be adapted in the future suddenly a cost issue in your mind? Keep in mind your conspicuous absence in the former reduces the weight of any possible answer to the latter.
How do you know the current commercial package can’t be fixed (or upgraded)? Where do you get your insider info?
Where was I when TriMet selected their existing trip planner? I don’t know. When did they select it? And when and where was that announced? Regardless, my lack of commentary on that has no bearing whatsoever on whether my questions about the cost (or any other aspect) of the new trip planner have any merit.
I can tell you, though, that if there was an existing, mature open source package when the old trip planner was selected, I certainly would have wanted TriMet to look at it along with any proprietary options and then select the best, most cost-effective option.
And the same thing is true now. If there’s a new platform that’s more cost effective and provides the same or better level of service, TriMet should definitely look into switching to it (assuming the cost of the switch itself isn’t prohibitive).
There are a lot of assertions about the new trip planner, but none of them are proven. How do we know that a majority of TriMet customers will ever find its extra functionality useful (where are the surveys to gauge customer interest)? How do we know it’s actually going to be more cost effective (where is the TCO analysis)? How do you know that it’s “better” than the old system (where are the side by side comparisons and results of user testing)?
Commercial software has a very poor track record for fixes. Or free upgrades. So it’s a very safe assumption that it’s not going to happen, and certainly not as cost effectively as putting one employee and a couple interns on the project.
This is why Portland can’t have nice things. Rather than going with a simple fix obviously full of win and letting it fly, someone’s gotta be critical and bury it in endless bureaucracy. I guess hater’s gon’ hate.
Your first statement is A) false (or perhaps just misleading) and B) easily applicable to open source software. It really depends on the particular software you’re talking about. I agree that there are certain advantages to the open source development model–the thing that bugs me is that you and others seem to believe it’s some kind of magic bullet and that it is *automatically* better than commercial alternatives.
As to free upgrades, it’s easy to argue that upgrades to open source software aren’t free either, especially across major versions. Even if there is no licensing cost, there is still a cost associated with upgrading–it’s always going to take some amount of developer time to perform upgrades. In addition, it’s often the case that companies (and public agencies) will purchase support contracts for open source software because they don’t have the time or expertise to fiddle with software that–you have to admit–is often poorly documented.
So, at this point, the question is whether or not developer time and support contracts are substantially cheaper than licensing fees and also worth the headache. For someone like you who is apparently dogmatic about open source, I guess that’s a pointless question, but you can’t assume that everyone else shares your values (and you can’t assert that they *should* either). For someone like me who doesn’t have a vested interest in this project, well, show me the numbers and let me make up my own mind about it.
Your “safe assumption” that the current software can’t or won’t be upgraded is really just a guess isn’t it? I’m guessing too, but I’m almost positive that there’s an upgrade path. Your statements about the relative costs and benefits are just a bunch of hand waving. Meanwhile, several questions remain unaddressed.
As to “Portland can’t have nice things”, what does that even mean? Portland has all kinds of nice things. There are several reasons for that, one of which is the fact that citizens ask questions and get involved. If this project or any other can’t hold up to public scrutiny, I doubt it was worthwhile in the first place. (Whatever happened to “transparency” anyway?)