OneBusAway Alert for C/D Line
Note that “8 minutes” is simply the headway divided by two.

RapidRide, notoriously, lacks a published schedule at all times at which the service operates at a frequency of every 15 minutes or better, which is from the start of service until 11 PM, every day. Moreover, OneBusAway has never worked reliably (if at all) for the C & D Lines (except for the late night publicly-scheduled trips). For myself, I know this lack of information makes me prefer the new Route 40 for most of the Ballard-Belltown trips I make, even though RapidRide serves the doorstep of my apartment building. While exchanging emails with Metro staff a few weeks ago, I inquired about these issues:

What is the rationale for not publishing a schedule for RapidRide in the periods where the bus runs less frequently than every 10 minutes? I’m aware of the given reason that it makes it easier to add buses, but it’s highly unlikely that Metro will ever need to add service in the off-peak on these routes, and 15 minutes is not, in fact, “so frequent you don’t need a schedule” (especially when the real-time arrival signs default to “refer to schedule” when one doesn’t exist). Moreover, it’s not possible to reliably plan a trip using the provided public data, as travel times vary so much throughout the day (particularly on the D Line), its not possible to know immediately how long your trip will take.

Alternatives already exist which provide peak-period operational flexibility to the agency while conveying to riders the off-peak timepoints required to plan reliable and fast connections; for example, Vancouver’s schedule layout, which is used for all of its bus routes:

Will Metro acknowledge the inadequacy of the current schedule information and commit to finding a format to better inform riders, for RapidRide and all frequent-service routes, as is done by other agencies across North America? Because no significant restructures will be happening next year, lots of staff time should be available to address this and other service-quality problems with RapidRide and the frequent transit network.

Since I wrote that email, Metro seems to have enabled real-time arrival signs at almost all the “station” stops outside of Downtown and Belltown (I mentioned why those stops were delayed previously), so many riders will know, once they get to a stop, how long they will wait. That is an improvement, but much of the point of a bus schedule, or OneBusAway realtime information, is to minimize time spent sitting in bus shelters, and it’s too late by the time you get there.

Metro’s response, and a discussion, after the jump. 

When developing RapidRide customer information we evaluated various methods of providing schedule information. We saw, similar to the example you provided for Vancouver, that when service runs every 15 minutes or less then it is frequent enough that exact schedule times are not necessary. The Vancouver information often gives a range of the headway (e.g. every 7 – 8 minutes) and not the exact times. This is similar information to what we provide, but in a different format. We have provided our riders information in the current format for two years now and we have seen riders on the A and B lines adjust to the change from our standard timetable format.

As we move toward implementation of the next two RapidRide lines, E and F lines, we will be reviewing how we provide line and schedule information for the full RapidRide network. We will likely make adjustments based on this review and input we receive from riders.

To specifically address transfer connections, if riders are making a transfer to other buses, then they can go into trip planner and put in an origin and destination and they will be given scheduled trip times.

A few reactions to this. First, I’m glad Metro has expressed an openness to making improvements to the current method of providing schedule information, based on feedback. As I’ve noted before, complaining en masse is a great way to effect minor service quality improvements at Metro, so next time you miss a connection thanks to RapidRide’s lack of schedule, don’t hesitate to complain to that effect: if enough people do it, something might happen. There are, however, a couple of respects in which this answer seems obviously flawed.

There’s a world of difference in the rider experience between eight versus 15 minutes of uncertainty in when your bus might show up. Saying the information presented is “similar” is perhaps true semantically, but the amount of usable information — when your bus should arrive, and how long your trip will take — in the Vancouver schedule is vastly more. It’s not clear how Metro’s evaluation “saw” that 15 minutes was an acceptable degree of uncertainty, but it flies in the face in the face of academic studies, Adam’s survey of similar transit routes worldwide, and the personal experience of everyone I know who has ever used a decent rapid transit or local bus system anywhere else in the developed world.

That Metro’s trip planner knows the unpublished schedule is good, I guess, but to the extent that information is accurate, it obviates any possible “operational flexibility” argument — even if that argument holds any water outside of the peak period (nobody outside of Metro seems to think so).

On the question of why OBA realtime information only works for late night “scheduled” trips, here’s what Metro has to say:

KCM’s project team is investigating this, and there may be multiple conditions causing these errors.

One element is a known issue with OBA. OBA has not correctly implemented software routines to properly handle headway based service. They’re working on this problem. As an interim measure KCM has provided a revised schedule to OBA which includes all scheduled RR trips. […]

KCM’s project team is working with the AVL system contractor to understand how the activation and deactivation of Headway Control affects the operation of the system and the data output that goes to OneBusAway.

Additionally, KCM does not always run with Headway Control enabled, in which case,unscheduled trips will never appear in OBA.

I’ve never seen this “all scheduled trips” data ever make it to OneBusAway (anyone else?). I’m glad to hear they’re working on fixing the problem, but it’s been six weeks now since the service change, which strikes me as a crazily long time to be working on this problem with no fix.

I understand Metro is overstretched and underfunded. I don’t expect miracles, but I do expect Metro to deliver this service at a level of quality that befits a premium, flagship service. The status quo of no schedule and no OneBusAway on the C & D Lines — more than six weeks after the start of service — is an intolerable embarrassment to the agency and is corroding the RapidRide brand. Someone at Metro needs to take charge of these issues an get them fixed: something immediate to make the service usable today, and and expedited reevaluation of the no-schedule policy in time for the February service change.

133 Replies to “RapidRide C & D: No Schedule, No OneBusAway”

  1. Also with the exception of late night trips there is no information given on OBA about the A line. no schedule no estimated times. it reports that no upcoming arrivals at the stop.

      1. Here’s an idea: The city of Seattle could install cameras focused on real-time signs at RapidRide stations, and Metro could monitor the cameras, and people could call Metro and ask for actual schedule information. Metro could also use the city cameras to write down actual arrivals, keep track of arrivals for a couple weeks maybe, enter into a spreadsheet thingy average arrival times at multiple locations, run off some copies on the Dittomaster machine, and mail them to all residents of King County. Metro could even put the live camera feeds from the city on the Inter-Net. Voila: schedules, real-time via computer/handheld, AND telephone real-time info!!

  2. Let’s not forget that Link — even more flagship, orders of magnitude more money spent — also has no published schedule and no real-time arrival. Thankfully OBA works, and the 10 minute headways cover up some of the sins. However, it was years before we got to even that point.

    1. There originally was a Link schedule, which they’ve since stopped publishing.

      ST & Metro: Please stop the idiocy and publish the schedules!

      1. Thanks Carl – it is a disservice and a condescending insult to passengers/customers for us to continue to be offered these “services” without a timetable. The corrections would seem easy, the lack of ease in planning connectivity makes this now carless reader more and more angry. Do I plan on an hour from Mad Park to the Admiral District, or 1.5 hours? Twohours ?? Two seats or 3???
        This simply cannot be that difficult a problem to solve.

      2. Agree with Carl & Lloyd – why are Metro so intent on alienating their most loyal riders and their fans? I was a big Metro fan and understand how difficult it is to jungle multiple service priorities. But not providing a schedule and blatantly ignoring the many requests for it seems like arrogance.

      1. The NYC subway schedules are in fact quite good. For example, the A train has two branches in Queens, one to Lefferts Blvd, and the other to Howard Beach (and normally to the Rockaways, that line is damaged.) Off-peak the A runs every 10 minutes and alternate trips continue on each branch. I find the A adheres to its schedule pretty well, and I use the schedule, rather than wait 20 minutes, especially when I connect to a bus in Queens.

  3. Ive found the trip planner and google maps will give times. I contacted metro about this in October and here was their reply:
    Hello and thank you for contacting King County Metro Transit.

    No, Rapid Ride lines do not have a set schedule. However, you can complete a trip plan that will give you an estimated time of arrival for your location.

    For help with trip planning and customer information, please contact Metro Transit’s Customer Information Office at (206) 553-3000 and press “0” to speak with a Specialist. You may also use the online trip planner at:

    Again, thank you for contacting Metro Transit.

  4. Metro has set times for departures from the terminals and on inbound trips there is a set schedule north of Market Street and south of Alaska Junction. But, southbound from Market to Westwood Village, every timepoint is marked “ET” and northbound, from Alaska Junction to Ballard, every timepoint is marked “ET”. So the first few miles of each run could have a published schedule and that would be great in the mornings when most riders are headed inbound. But it doesn’t make the afternoon or evening commutes easier or more reliable.

    If the RR buses terminated in downtown, there would be better overall schedule reliability but that would also raise the cost providing the service. It looks like the E Line will terminate and turnback in downtown Seattle. I wonder if there will be any way to integrate the C, D and E schedules to provide better reliability on the C and D?

    1. Metro has been providing timetables with “ET”‘s for years. This would be much better than no timetable.

  5. So there is a schedule, they just don’t want us to see it. But if you call, they’ll tell you. Gee, that’s nice.
    There’s a lot of things Metro does wrong that has nothing to do with money. This is one of them.
    As a rider in an area with frequent service (Belltown) I’ve found this route basically worthless. If my choices are to stand at Cedar St waiting for who knows how long for a D to show up, or walk to Blanchard and catch a 40 that I know is 8 minutes away, I’m going to walk every time.
    Maybe that’s what they want, I don’t know. But at this point it almost seems like they’re trying make it worse than it needs to be.

    1. Just publish the schedule

      People who don’t care can ignore it. People who like to plan will use it.

      Withholding the schedule while operating to a schedule is the height of bureaucratic arrogance, and is the kind of thing that makes people hate monopolies

  6. I’m just going to toss out the observation that every time I’ve narrowly missed a bus at Leary lately, the real-time sign has defaulted from “due” to “18 minutes”.

    So many lies behind RapidRide. Is Metro lying about 15-minute headways too?

    1. No. The internal schedule has 15-minute headways in the periods when indicated on the public “schedule.”

    2. d.p., It’s just inconsistent headways in operation. The other day I was going to LQA in the morning, and the sign changed from “Due” to “5 Minutes” when I just missed the bus. So the opposite happens as well.

      1. Except that Leary is less than 15 minutes from the terminal. So the real-time predictor is going off something other than actual location.

      2. @d.p. Metro has been having issues with the GPS system that Init is installing on the bus. From my understanding one of the issues is how the real-time arrival system deals with buses when they are laying over at a terminal. I’ve heard this from VeloBusDriver with relation to the B-line.

      3. if you are near a terminal weird things can happen because it will think the bus leaves 15 minutes early so it shows next bus in 29 minutes when in fact you only have 14.

    3. The same thing happens on RapidRide B: the real-time signs have implausably long wait times, like 29 minutes during a 15-minute period, or 13 minutes during a 10-minute period. But the buses come within the expected 10-15 minutes anyway.

      1. If they’re that wrong they should just turn the signs off. Why are they having so many problems with such standard technology?

      2. Complete guess: These are at the beginning of the line. It has to be tricky to figure out how late the bus is before it’s finished it’s previous run, especially if the driver’s allowed to use the restroom (for example).

      3. They are at the beginning and the middle of the line. At 124th (before the complete 124th outage), they sometimes showed 13 or 29 minutes as I said. At Bellevue Transit Center, they seem to be timed for the second or third bus that hasn’t arrived yet, not the first bus that’s standing there.

      4. especially if the driver’s allowed to use the restroom (for example).

        And that’s why RapidRide is no “high-capacity transit corridor” in any regard.

        Imagine a London Underground train leaving a terminus 5 minutes late because the driver is in the loo!

        If your vehicles are still prone to such incredible and unpredictable delays that the bathroom break at the end of the run is in peril, then either:
        – Your bottlenecks are not sufficiently addressed;
        – Your service isn’t frequent enough to even out demand fluctuations;
        – You don’t have enough extra operators running the route to ensure sufficient recovery time.

        In the case of RapidRide, it’s all three.

        Late vehicles on any “core” service because of the bathroom. Yeesh!

      5. There shouldn’t be too many restroom issues on C/D. It’s actually scheduled for very good recovery time by modern Metro standards.

      6. “There shouldn’t be too many restroom issues on C/D. It’s actually scheduled for very good recovery time by modern Metro standards”

        Don’t just look at time, look at distance/access to a toilet. I could write a book on the impacts of poor comfort station access on system reliability. Many routes have no comfort station within a 5 minute walk (walk 5, pee/wash, walk 5). The 1, 245, and 73 come to mind but I’m sure there are plenty of others. The 54/15 was horrible as you’d drive all the way from West Seattle to Crown Hill, get a short break (if lucky) to access a comfort station inside a locked church that was a short walk from the terminal. I have no doubt the C/D terminals are better, but that all depends on the arrangements Metro has made with area businesses and/or whether they have built a dedicated comfort station.

      7. The comfort station situation is good too. On the C end you have the Westwood Target within a couple minutes’ walk. On the D end you are right outside the QFC.

        I didn’t know the 54/15 were ever through-routed. When I drove the 15, the QFC was still there at Crown Hill. I’d leave the bus with flashers on in the 15th/85th NB zone and just go into the QFC.

        The route I remember as being horrible from a comfort station perspective was indeed the 1/36 (when the 36 turned at Beacon & Myrtle). I’d ask the coordinator to drive up to the 2 terminal to use the comfort station there, and they’d usually let me.

        Another bad one was the old 14 before MBTC.

    4. Really? RR buses are scheduled every 10 minutes from 4-7 at all points along the C and D? I doubt it.

      1. Yes, they are (at least in the peak direction). But it doesn’t feel like it because they have been having serious reliability issues inbound from Ballard to downtown. It seems to be a combination of overloads/extremely high ridership, overly ambitious scheduling, and poorer than historical traffic conditions possibly exacerbated by weather.

      2. The “schedule” says every 10 minutes. it doesn’t say “in the peak direction”. Sounds like a major “stretching of the truth”. Do you know if they pull the same fast one during the supposedly 11 p.m. transition to 30 minute frequency? The more facts that come out, the worse Metro is making themselves look.

  7. Why don’t people just use the point-to-point trip planner? If you set it for all trips between 5 AM and 4 AM, it provides a complete schedule.

    1. Trip planners are antithetical to the very idea of frequent, easy to understand core transit service.

    2. I just tried that. 5 am to 4 am, weekdays, from 15th & Market to 3rd & Pine. This is the entirety of the resulting schedule:

      05:07 AM 05:30 AM MT 674 D LINE TO DOWNTOWN SEATTLE (Ends as MT 673)
      05:37 AM 05:45 AM MT 674 D LINE TO DOWNTOWN SEATTLE (Ends as MT 673)
      02:44 PM 03:10 PM MT 674 D LINE TO DOWNTOWN SEATTLE (Ends as MT 673)

      Three runs, eh? So much for that complete schedule.

    3. Because many people don’t know it exists and it’s an unintuitive way to get basic information that should be on the main schedule page for RapidRide (and Link).

      1. I have. I’ve written articles on this blog about the lack of Link schedules. I’ve even created an unofficial schedule for people to print out and mock ups of at-station schedules.

        Metro already told people to use Trip Planner and it’s a poor solution to answering a question as simple as “when’s the bus/train coming?” This is a basic responsibility in providing information that shouldn’t be left to private individuals.

        The ball is completely in Metro and Sound Transit’s court. Give the riders what they want.

      2. Oran, when are you going to publish a RapidRide schedule? Sounds like that’s what the people want!

  8. If google maps and several other apps can provide schedules, why can’t METRO? I printed a schedule for the Alaska Junction inbound RR and it was accurate nearly to the minute. I have not missed a bus since. METRO is full of IT!

      1. I apologize, I thought I had found this on google transit. Basically, work your way through the menus,down to the stop level and the schedules are there. Real time info is not. However, Inbound from Alaska Junction has worked for me every time since I started using this site (whis is also an android app) Outbound from DT is basically useless due to traffic and other factors.

  9. I agree with you about the 40 having more arrival information via OBA, however, not sure that information helps as the 40 going north through SLU is often FULL and doesn’t even stop…I have had northbound 40’s just drive on by 4 times in the last three weeks. Seems like this is a route that deserves more service as well…

    1. I was wondering when this would happen. I assume this is from more and more people giving up on RR-D and switching to the 40?

      1. When I’ve ridden the 40 lately (very early peak) downtown demand has been lightish, with SLU where the demand really starts.

  10. I continue to suspect that the real reason behind the resistance to providing a schedule is that a published schedule, compared to reality, will reveal how utterly unreliable these buses actually are. If they think this disinformation is doing anything to protect the RR brand, they’re obviously pretty profoundly wrong about that, but no other explanation they’re giving makes any sense. Comparing to routes with actual 8 minute headways is the real tell. Metro doesn’t have those, but it does have lots of non RR routes with similar headways. (I’m familiar w/ the 5, 44, 48, and 358. I’m sure there are others). For some strange reason, the apparently complete lack of utility for schedules with 10 peak/15 day off-peak headways hasn’t inspired them to bring this grand innovation to other routes.

    WHen I sent my complaint email about the lack of schedules/OBA, I specifically asked why they’re continuing to publish schedules for buses with the same headways. That part of the question was ignored in the boilerplate response I received.

    I’m really quite understanding about most of Metro’s many flaws. Overstretched, understaffed, and trying to provide reliable transit with the kind of traffic Seattle deals with is a brutal place to be. But this particular flaw makes me feel, unlike others, like Metro is treating me like a dim-witted mark, rather than actually trying to figure out how to best serve me given existing constraints. I’m honestly a bit surprised by how much it pisses me off.

    1. “I continue to suspect that the real reason behind the resistance to providing a schedule is that a published schedule, compared to reality, will reveal how utterly unreliable these buses actually are.”

      This is exactly why ST stopped providing a published schedule for Central Link — Link was embarassingly unreliable.

      Is there a schdule for the S.L.U.T.? I don’t remember every seeing one. If not, why not? With a route only a little over 1 mile long, it should not have any trouble keeping to a schedule. Or, does it?

      1. In any case, Link’s unreliability was a problem in the first year, but I’ve never had to wait longer than the expected time since then, and sometimes the headways are pleasantly shorter than advertised, particulary around 7 and 8pm. The only time Link seems to be delayed is when there’s an obvious outage in the system; e.g., an accident on the tracks, a bus breakdown in the DSTT, or Obama visiting the Paramount.

      2. When Link opened, ST had a chart showing the travel time between each pair of stops. That chart has disappeared (except for the Wikipedia page). One must use a trip planner to discover that Westlake to Airport is scheduled for 38 minutes.

      3. With Link, travel times are roughly the same throughout the day, so this page can also serve as somewhat of a guide.

        I think the horizontal strip for headways is the wrong format; it should be vertical. But keep the vertical strip to show travel times between stations.

      4. I always assumed travel times were the same throughout the day, but it’s several minutes shorter in the evening. One evening I went from Mt Baker to Westlake in a speedy ten minutes (14 on the schedule). Intrigued, I timed Othello-Westlake another evening: less than 20 minutes instead of 22. I also timed Intl-Dist to Westlake. In the daytime I normally get 10 minutes for that, but that evening it was a lightning fast 3 minutes, and 1 1/2 of those minutes were delays that should go away when buses are kicked out of the tunnel.

      5. Link doesn’t keep to a schedule but they do try to maintain specific headways. You can see the clocks that are used to control headways at Rainier Beach (northbound) and Mt. Baker (southbound). They are at the end of the platforms. It would be nice if other stations had the clocks, too. No more waiting for the 2 minute warning announcements.

    2. They can do schedules for the other high(er) frequency routes because the buses aren’t red. Everyone knows you can’t have a schedule with red buses.

      1. Sydney seems to have exactly the same problem with their Metrobus service.

        Regular routes served by blue buses run at 10 min frequencies all day and have a printed timetable. But the Metrobuses run every 15 minutes and don’t have a timetable simply because they are red.

        At least the red buses in Sydney require that you pay before you get on for all of the stops.

  11. but it flies in the face in the face of academic studies, Adam’s survey of similar transit routes worldwide, and the personal experience of everyone I know who has ever used a decent rapid transit or local bus system anywhere else in the developed world.

    Well, you know, the hills and the soil and the

    1. It’s the unusual, highly consolidated subglacial soils.

      (Oh, no, that’s why a Deep Bore Tunnel was a mistake. My bad.)

  12. For myself, I know this lack of information makes me prefer the new Route 40 for most of the Ballard-Belltown trips I make, even though RapidRide serves the doorstep of my apartment building.

    Same here, though with the 28. It’s a couple blocks farther than the D and comes half as frequently, but it has a schedule and accessible real-time arrival information, so I can time leaving the house to minimize waiting in the rain, I can figure out what transfers I can or can’t make downtown, etc. With the D, I could end up waiting 2 minutes or 20; maybe I’ll be able to transfer to an 8, maybe I won’t. It’s just not worth the bother.

    1. Wow, someone should blow up this comment and put it on a billboard in view of Metro headquarters.

      I remember reading the discussions on here before the new RapidRide routes came online. Many people were suggesting that Metro consider eliminating service from 8th and 24th Avenues, since RapidRide would be so awesome that people living near routes 28 and 18 (now 40) would be better served by walking a half mile to RapidRide’s so-frequent-that-you-don’t-need-a-schedule service than waiting around for their existing 15-30 minute buses.

      How funny that the exact opposite is happening — at least one person who lives near the D line is admitting to walking farther to the 28 because you at least know when it’s going to show up.

      1. You wouldn’t just delete the 28 and 18 in isolation, you’d fold the service hours into the D so that it would be more frequent and reliable. And it was really about the 28 and 17, not the 28 and 18. I had assumed deleting the 18 was part of the regular plan for the D because it would need the 18’s service hours to get to the 15-minute frequency threshold. So I was stunned when the 18 was retained as the 40 and actually increased frequency daytime.

      2. No-one ever planned to delete either the 28 or 18 without some sort of replacement — route spacing that wide would lose a ton of riders.

      3. Mike,

        Before October, the all-day frequencies looked like this:

        – 15: 3 buses an hour
        – 17: 2 buses an hour
        – 18: 3 buses an hour

        Now, they look like this:

        – D: 4 buses an hour
        – 40: 4 buses an hour

        So, in fact, the change did move service hours from Leary/24th to 15th.

      4. Regarding the 28, Metro only suggested not serving 8th Avenue south of Market Street, because the 40 would provide service along Leary. Right now, with having both the 28 and 40 there’s actually a lot of service in Frelard.

      5. That’s good to hear. The 28 actually seems to have a lot of off-peak traffic, at least when I ride it, mostly between 100th and Leary. It’s not uncommon for late morning buses to be SRO by Fremont.

      6. Metro never suggested deleting the 28, but it and the 17 and the 26 have been in my target crosshairs. They are low-frequency residential routes with a lot of single-family houses, and each is within a 10 or 15 minute walk of more urbanistic routes that should be more frequent. We could have a 10-minute frequency standard on the trunk routes if we weren’t running all these 30-minute routes parallel to them.

      7. I agree (and have always agreed) with Mike about route consolidation.

        But after the way Metro fucked Ballard on both RapidRide frequency and 40 evening/Sunday frequency — especially the glorious middle finger of cutting total late-evening service in half — I simply don’t trust them to implement service consolidation at the level necessary to make it worth the extra walk.

        Half-ass your restructure once, shame on you….

      8. The 28 is not comparable to the 17 or the 26. It’s well-used, and covers an area out to 3rd Ave NW, way out of walking distance from the D Line.

      9. Before October, Interbay and uptown Ballard frequency looked like this:
        Peak 10
        Midday 10
        Night 15
        Sat 10

        And there was a printed schedule. I now must wait an average of 7.5 minutes during the non-peak and Saturdays, where before I waited for about 2 minutes, with some level of certainty. Even if I couldn’t find a schedule, my average wait was 5 minutes. Where is Metro?

  13. “To specifically address transfer connections, if riders are making a transfer to other buses, then they can go into trip planner and put in an origin and destination and they will be given scheduled trip times”

    I have found trip planning for connecting service to the B Line to be unreliable at best. Trip planner will often send me in the wrong direction, completely avoiding the faster B Line, to connect to other service with a set schedule. The resulting trips are much longer and don’t make much sense. I wrote about it last year. This particular issue for the time in that post has been fixed, but Google Maps and Trip Planner both give some pretty wonky trip plans.

    Every time I see a confused passenger holding a trip planner print out, I take great pity on them and do the best I can to get them where they need to go. Trip planner is a useful tool but needs to have a big disclaimer attached to it for many trips. I report every problem I see with it and Metro fixes it, but it has the feel of holding down a balloon only to have it bulge out elsewhere.

      1. “Not to mention that the mobile site for ST’s Trip Planner is actually good”

        Except that it doesn’t recognize Chrome on my Nexus so I get the desktop version instead :(

    1. I’m trying to use trip planner for the two end segments of the trip and assigning an arbitrary number of minutes to the Rapid Ride segment based on headway and what I think Metro uses as point to point time (although I wasn’t able to find that on the web site except for the early morning/late night times).
      I should avail myself of the opportunity that Metro is giving me to explore the city using the route they suggest in Trip Planner.

  14. If all these real-time issues get sorted out, another useful bit of information to provide would be the current state of service of frequent bus/train lines. For example: “Good Service”, “Minor/Major Delays”, “Suspended”, “Works Alert”, “Travel Alert” (see Metro Melbourne’s site). A simple color code would tell people what to expect.

    PS: Why the Metro Melbourne example? Because of this wonderful PSA about safety in mass transit.

    1. Wow. In the US no station would allow PSAs that *long* to be aired, which is why such cool things only come from Australia etc…

  15. Just to let you all know, the situation with the real-time feed for RapidRide is a bit more complex than has been described in the article. OBA and KCM have been going back and forth to try to get a valid feed that works for RapidRide C & D and there obviously have been delays with the other fixes needed resulting from the 9/29 change along with the 11/5 change.

    The workflow for data from a highlevel can be described as: Beacon and now GPS data received by Metro from buses, and is incorporated in a real time stream that provides estimated ETA data. This data is fed to a small server that then converts it to GTFS for OBA. OBA (as I understand it) does not do any predictive computing but rather reports what is in the feed for each stop.

    FYI, both Will and I are “OneBusAway Ambassadors” and have been assisting with data collection and system observations. For example, of late, we’ve been tracking GPS stop announcement failures. If you use OBA, we encourage you to provide useful feedback including service issues. We forward relevant info back to the agencies.

      1. I can’t tie my shoes without violating a patent. The patent system is largely beyond its usefulness in encouraging innovation… which, ironically, is why it continues to exist.

      2. Patents on computation are invalid. All of them. They violate basic principles of patentability.

      1. From how its been explained to me, It basically comes down to how the routes are programmed into the GTFS that forms the base data layer for OBA. RapidRide A/B were programmed like any other bus route – there may not be a public schedule, but in the GTFS there is one, and so everything works like it should. RapidRide C/D, however, use the new Headway Control feature, which changes both how those routes are programmed in the GTFS, and how their real time ‘pings’ are sent to OneBusAway. OBA has never seen data like that before, and has no idea what to do with it…so it just shows the headway divided by two. This, by the way, explains why NiteOwl trips work fine: they’re scheduled in the GTFS like ‘normal’ routes.

        Metro did get us a patched GTFS where C/D are programmed like A/B, but doing so required a lot of tweaking on their part, and combined with all the schedule tweaks since 9/29 (and 9/29 itself!), getting things to work has been difficult. It is in progress, though, and from what I’ve heard it’s close to being fixed.

  16. If people want Metro to print schedules, which would seem to make sense for a transportation service, they shouldn’t bother contacting Metro. Sometimes, Metro needs to be told what to do. Contact Executive Constantine or councilmembers McDermott or Phillips.

  17. I hope nobody ever forgets this Rapid Ride debacle when the BRT cheerleaders hail it as the “equal but cheaper” alternative to rail.

    1. Being as Link doesn’t have a schedule either, I wouldn’t get too carried away with that line of argument.

      1. I’m speaking more in generalities…the (lack of written) schedule problem is not a huge structural flaw, IMHO.

      2. Additionally, OBA works like a charm for near term scheduling…so I think it’s absence from (3 out of 4) RR lines is a much worse offense. OBA has replaced the old fashioned schedule, in most cases I bet…well, at least from my observation of what people are using at Capitol Hill and UW area bus stops (can anyone really argue that for, say the 43, a schedule out performs OBA?).

      3. But Link seems to run fast enough and reliably enough that people aren’t crying out for schedules…. which says something.

  18. I am baffled- if the buses have GPS beacons, why wouldn’t somebody just offer a live map with the each numbered bus showing its current location? (Maybe because you’d have to cross-reference a list of what bus route that particular bus is assigned to???)

    1. GPS and beacons are two different things. Right now, Metro doesn’t give anyone access to the GPS data except OBA. The system isn’t finalized, so put down the pitchfork.

      Also, knowing the current location isn’t all that useful. The useful part is guessing when it will get to another location knowing its current location. Figuring that out is patented.

      1. “Also, knowing the current location isn’t all that useful.”

        I’d pretty strongly disagree with that. It’s not at all hard for a rider to know when to show up at a stop based on a vehicle’s current location. Used to do it all the time with Chicago’s real-time map (available for every bus and train in the system), and still do it when I occasionally ride the SLUS.

        In an ideal world, yes, both should be available. But since it is apparently incredibly complicated for Metro to provide real-time arrival information, they should just open up the date and let anyone interested develop a real-time map.

      2. That’s actually what MTA’s “BusTime”‘system does: it’s shows a dot on a map. Interestingly, “BusTime” is actually OneBusAway, with a some modifications and improvements.

        As for open data, the data is open. Metro’s static GTFS is public – you could download it now if you’d like. OBA’s data, including real time, is available through an open API that multiple services and applications use. Eventually, everyone would like to see it all available direct from the source and in the same format, but that takes time and money that neither Metro nor OBA (which is run by UW in partnership with Georgia Tech and local agencies) have. SoundTransit is doing some exciting things with their R&T program, though, and I think that will be the ultimate solution.

      3. Tim- No pitchfork here. I reserve my pitchforks for planners talking about a Lynnwood LINK instead of digging tunnels or erecting gondola pylons in West Seattle. :D

        Oran- Dang, that would be EXACTLY what I was looking for- thanks. Now why aren’t more people using this to get an idea of where their bus is?

  19. I think the issue here is something more basic. There’s a strong desire among transit planners and advocates to move towards headway-based services, and a more transfer-oriented network. But (in my opinion) the headway where that works is 10 minutes or less, not 15 – and nobody thinks we can afford ten minute headways. At fifteen minute headways you need a schedule – or at least a clock-face schedule (where signs tell you buses come at :10, :25, :40 and :55 after the hour, or whatever).

    It’s important to get more precise about what’s really needed to change the transit paradigm. I think 10 minute frequencies are needed before you can throw away timetables and tell people transfers are always OK – and I think the comments here support that. If 10 minutes is the threshold, then you need a different strategy if you can’t afford that (like clock-faced headways and timed meets). Or you need to find a way to prioritize 10 minute service on headway-managed routes and transfer points and reduce service on other routes. (And if 10 minutes *is* the threshold, you need to spend a whole lot more on headway management service control to avoid the bunching that’s bound to happen on Seattle’s traffic-calmed streets). But pretending that a 15 minute headway is enough to change the rules won’t work, in my opinion.

  20. I Was a big fan of metro and applaud their efforts to provide the best service they can with limited resources, and although I don’t agree with all their routing and service decisions I can understand why they make those decisions.

    But their attitude to providing schedules for RR just makes me angry (as a rider) and sad (as someone that desperately wants to support Metro). This is made worse be the unwillingness to compromise on this issue which is vitally important to many of their riders. We need a schedule. Metro, just go ahead and publish one. You are doing it already in the trip planner, so just make that info available in a more more accessible format.

    What is truly depressing is pointing to RR A and B as examples of where lack of schedules is successful, and implying that nothing will change until at least RR E/F start.

  21. Dear Metro: Spend 15 minutes standing at a shelter in 40mph gusts in a downpour like today’s, then tell me that we don’t need a schedule for the D.

  22. Incorrect information is worse than no information. With on-time performance like Metro’s, a schedule could do more harm than good. Priority 1 should be getting real-time data functioning properly.

    1. At least during the 15 minute headways, it would be nice to have a schedule. Reliability during the late evening, midday, Saturdays, or Sundays is not really that bad.

    2. Depends. Incorrect real-time information is worse than scheduled information but schedule information is still better than no information. So I agree getting the real-time information working should be priority number 1 but schedules should still be provided at least for 15 minute headways for those that don’t have access to real-time info or are planning a trip ahead of time.

  23. It’s official – the role of “rapid ride” is to grab people who were planning on waiting for the 40, but happened to see it come first. Looking at it that way, it does its job pretty well.

    1. Except the 40 and the D Line share only three stops:
      -NB at 3rd/Virginia
      -SB at 3rd/Virginia
      -SB at Holman/Mary

    2. Sort of,but when heading downtown from Central Ballard on the 40 you have no way to know whether to get off the 40 and switch to the D at a “rapid ride” stop, since it’s impossible to know when the D might show up. Again, how frustrating to not have a “rapid ride” schedule. The folks who think 15 min is frequent enough to not need a schedule are obviously not transit riders and don’t place a high enough value on transit riders’ time — i find that demeaning.

      1. Of course you don’t get off the 40 to switch to the D – even if the D is right there. But if you’re waiting at a common stop and happen to see the D come by first, then getting on the D makes sense.

    1. See Will Green’s comment. Link is just schedule information that I think has been reversed engineered.

Comments are closed.