After years of protests that it couldn’t be done, and six years of study and work, Sound Transit has finally found a way to put next train arrival times on the existing message boards at stations that opened in 2009. Capitol Hill and UW Station opened with this capability in 2016. Between Angle Lake and Westlake, arrival times rapidly rotate with other messages. Riders need wait no more than a few seconds to get the key info.

The picture above is on the mezzanine level. There is a similar sign on the tunnel entrance, a good indicator on whether or not to hustle. At the platform level, the signs only display the relevant direction (see below).

From the agency’s blog post:

In addition, Sound Transit will have the ability to interrupt real-time arrival information to update riders in the event of a service disruption.

This allows for the signs to have more relevant information during those times by displaying specific messages until full service is restored.

This feature can be activated for the entire alignment at once, or at individual stations or groups of stations as needed.

By email, ST’s John Gallagher said that though they might tweak the format, this is pretty much how it’s going to look until ST revamps the whole messaging system in 2023. As far as I’m concerned, now that we have this, they can take their sweet time in shipping a new solution.

58 Replies to “Next train signs are finally here”

  1. Why is it so damn hard to say northbound and southbound? Why does no transit agency include this incredibly useful information?

    Follow the example of highway departments and include direction AND major destination. Looking at the photo, there is obviously room to show both.

    1. What’s wrong with using the line termini to indicate direction?

      Using directionals can lead to confusion. I-90 for example runs east-west, but in the Chicago metropolitan area it runs generally north-south and remains signed eastbound/westbound. East Link is similar. Is an East Link train departing downtown Bellevue towards South Bellevue a southbound train or an eastbound one?

      1. That should be *westbound. My mistake.

        Anyway, it would lead to confusion for people unfamiliar with the system.

        “huh, you want me to take a westbound train to go south?”

      2. “What’s wrong with using the line termini to indicate direction?”
        1. Sometimes the terminus is in a backwoods place where no one is going, especially on bus routes. Even on Link, Angle Lake is not a major destination, nor would you expect visitors to Tacoma to know where the Theater District is.
        2. It requires all of your riders to know where that terminus actually is, a problem for visitors and infrequent riders. Visitors will all know the airport, but they won’t know Angle Lake, and they might not know where UW is. Even locals who ride infrequently might not know where Angle Lake is, for that matter, since the term wasn’t commonly used for that location before the station.
        3. Cardinal directions are something that you can instantly see on a map. A terminus (or a reference station) is something that people have to spend time looking for. A person can catch a train labeled “south”, without looking at a map, and know that they’re going the correct direction, which is not true of a train labeled “Seatac”
        4. If any transit line can’t easily be described as north/south or east/west, then the problem is the routing, not how you post the signs.

        As far as directions, the answer is to do what highways do and use the “big picture” direction. Eastlink would be posted east/west for the entire line, even if the stretch in Redmond goes north for a bit. This won’t confuse anybody because highway 520 already does the exact same thing in the exact same place, and everyone knows that Redmond is east of Seattle.

        Of course, your response ignores my central point to include direction AND destination.

      3. Terminals are extremely obvious on the line schematics posted in trains and at stations. Schematics are easier to deal with than maps. They’re a straight line with points on it. People (especially people unfamiliar with a system) mostly don’t get on trains with a cardinal direction in mind, they just know what stop they need to get off at. It doesn’t matter at all if you know where Angle Lake, Azusa, Malden, or Inwood-207th are geographically if you can see them on the schematic relative to your destination.

        >If any transit line can’t easily be described as north/south or east/west, then the problem is the routing, not how you post the signs.

        What of extremely common U- or L- shaped routes?

        The current system just works. I’ve never heard of anyone accidentally getting on a train going the wrong direction because they don’t know where Angle Lake is. There’s no need to add extraneous information to the signs that will just confuse people.

      4. Donde, pretty much every train system on the planet uses the terminal station to indicate line direction. I see no reason for ST to reinvent the wheel (again) especially when it adds more unnecessary noise. They’ve finally been convinced appending the word ‘station’ to the end of every stop is redundant.

      5. If the correct line and terminating station aren’t obvious on the map, then the map needs to be redesigned. If there’s not a map right there to consult, then they need to put up more maps.

        Most networks identify lines by color, name, or number, and show direction by the last station. This also works for trains going only partway, which show their last station.

      6. Except London, none of the largest and busiest metro systems around the world use cardinal directions to indicate train directions. It makes more sense for buses because they use the road network but even that is only applicable if your city and routes are laid out on a grid.

      7. London only uses EB/WB/NB/SB for indicating platforms, not on the trains themselves which follow the “[Line] to [Destination]” convention. Some platforms don’t even have a direction assigned, just a number.

        The Circle Line isn’t actually a continuous loop anymore. Since it shares all of its stations with other lines, those platforms follow the direction of that line. As for announcements and signs on the train, they use “Circle Line via [1-2 key stations]” to indicate direction.

  2. Please tell me that sound transit isn’t:
    1) Implementing this by paying a full-time employee to sit in a control booth and manually type in the information from OneBusAway every 30 seconds.
    2) Having the obnoxious tunnel voice scream arrival times into our ears every 30 seconds, in addition to all the other spammy announcements?

  3. I’ve also noticed one train that announced “this is a train to angle lake station” at every stop last week. Is this prep for adding East link or was it a glitch?

    1. I think that’s just a driver preference. I’ve been on a couple of trains that do that at every station, and I think the driver was manually having that announcement go at every stop.

      1. Sounds plausible — this was the rationale for expanding the audible bus announcements from major stops to all stops, that it was an accessibility thing.

  4. My god, you mean the signs didn’t need to be upgraded, as claimed numerous times by Sound Transit? We all knew that was a bogus excuse.

    1. We don’t know what it required behind the scenes, but ST said it was expensive and it didn’t have the money until now. It was part of the ST2 budget. The cost may be reasonable or it may have been overcharging by one of the few companies who work on these older systems, who knows.

      1. Mike, you put a lot of good faith into Sound Transit when “we don’t know what it required behind the scenes” is exactly the problem. We’ve been asking for a decade why real-time arrival isn’t there and they’ve done nothing but begrudgingly issue provably false excuses. Now suddenly it appears, I’m happy this is behind us, but now I just want to know why it took so long and why our questions weren’t taken seriously for all these years.

  5. Sound Transit really seems to struggle with some of the basics: arrival times and working escalators.

    They seem to be slowly learning over time, but I wonder don’t they have people on staff who have worked on other major transit systems like BART or DC Metro? Not everyone at the agency should be learning on the job. There should be some veterans in the room who can say, “no this won’t work,” or “we need to take care of this.”

  6. This is a huge improvement and should be celebrated. Why all the negativity?? Comments sections seem like they are getting more cynical and angry every day.

    Yes, it isn’t perfect and should have been done years ago but it is done now and hopefully it will be successful. I was very excited to see this on my trip to work this morning…

    1. There’s negativity because 3 years ago Sound Transit said exactly this:

      “The existing ‘dynamic’ screens are severely limited and cannot accept the heavily processed data feed that OBA uses. The study estimated the total cost of replacing the screens at $475,000, which ST deemed ‘prohibitive.'”

      This, as it turns out, was either a wildly misinformed excuse or it was a lie. Some of us knew we could give a high schooler one weekend with a Raspberry Pi and they’d figure out a way to get RTA displayed on those screens.

      1. Well I don’t think it was a lie, but it was a nonsense answer from a technical perspective. The screen can’t process the “data feed,” but I think they meant the OBA user interface. You obviously can’t show a web browser on those dot screens, so they have to make a new UI to translate the feed data to the screens.

        The fact that they referenced upgrades to the signs to be like the ones at CHS and UW (which are just Windows 7 PCs) seems to indicate that they were walking about the OBA website when they said “data feed.”

        Clearly they were working on a technical solution so I wish they would speak in correct technical terms.

      2. ST didn’t say it was prohibitive forever, it said it was prohibitive right then. ST2 included a project to do just that, and it was done at the scheduled time.

      3. No, Mike. ST said it was prohibitive because the LED signs were incapable of displaying real-time arrival info and would have to be replaced with new LCD signs. That’s obviously not the case because here they are, displaying the info they said was impossible.

      4. I think you’re overinterpreting it. LEDs can obviously display any combination of pixels, but it requires software to drive them, and the units may have limited memory and and the feed lines limited bandwidth. ATMs still use 300 bps modems even after home modems reached 56,000 bps because that’s all they needed for their original purpose. So the bottleneck was clearly the software, and maybe the hardware limitations required them to design a custom translator between the data feed and the signs. Or maybe ST thought thought it was cost-prohibitive then but later found a way to do it. In that case we should be glad it found a cheaper way.

      5. Mike, they very specifically said the screens were too limited and would need to be replaced. I’m not over-interpreting anything, that’s the excuse they gave, which we all know doesn’t make sense from a technical perspective. If the issue were indeed software they could’ve said that but they didn’t.

        I’d be interested to know how they suddenly solved this mysterious problem. Until then, I’ll assume they actually had no idea and didn’t think it was a priority.

  7. Does anyone know if the future East Link light rail line will go from the eastside and then end at the Chinatown/ID station? Or will it continue north through downtown?

    For example, would a person be able to ride from Bellevue to Westlake without transferring trains?

    1. It will continue north along the entire line, to UW, Northgate, and eventually Lynnwood. No transfer necessary.

    2. The current plan terminates at Lynnwood full time. (The earliest plan terminated at Northgate off-peak.) There’s no turnback between Stadium and Northgate, so all trains that reach Intl Dist will have to continue to Northgate at least.

      When the Everett extension is built, the blue line (East Link) will terminate at 128th in south Everett, while the red line (from West Seattle) will go all the way to Everett Station.

      Of course, ST may change its mind later depending on ridership. I could see East Link terminating at any of Northgate, Lynnwood, Ash Way, 128th, or Everett Station (unlikely) in the lowest-use periods.

      1. After Northgate, what’s going to happen to the turnback track at UW station? Are they actually going to remove it?

        I hope they don’t, because it would be senseless to operational flexibility that is already built. Otherwise, a disruption in the ULink tunnel would halt Link service as far north as Northgate.

      2. I’m not sure where Mike is getting this information. There’s a turn-back after Westlake (because that was the terminus before U-Link) and ST is building a turn-back at IDS for East Link. So if we include the one after UWS there are several (3) turn-backs between Stadium and Northgate.

      3. There’s a crossover after Westlake and one before UW station, but don’t confuse those with a turnback track. To turn a train back the driver needs a place to stop the train, lock down that drive station, leave the operator’s cab, walk to the other end of the train insert key, or whatever credentials they need to engage the drive system. Without a dedicated turnback track (think of the third track south of Stadium and Rainier Beach) the train would be blocking one direction while this happens. It may be possible with careful timing to do this when trains run only every 8-10 minutes, but once combined frequency is every 3 minutes it’s impossible.

      4. I could see East Link terminating at any of Northgate, Lynnwood, Ash Way, 128th, or Everett Station (unlikely) in the lowest-use periods.

        Anyone want to take any bets? If I was a bookie (and I was, briefly, in high school) I would set the odds like so:

        Everett 5-1
        128th 3-1
        Ash Way 3-1
        Lynnwood 2-1
        Northgate 2-1

      5. Mike is correct about the current ST plans. I’ll just add that drivers need breaks and these lines are pretty long as light rail lines go. The decisions on where to turn back trains are based on driver work rules as well as where a siding track is.

        I’ve witnessed turning around light rail trains in the middle of a line without a siding in San Francisco, and the time it took to do that really created backups. It’s easier at end stations like UW and Angle Lake because the reversing time can also be boarding or break time, and each of the two tracks at the station doubles as a siding.

  8. This is great! I noticed it yesterday morning and it immediately helped me realize that I didn’t need to rush down to the platform to catch the train. The arrival estimates seemed accurate at Westlake.

    They also changed the screens at UW and Capitol Hill to a 2-column format–presumably in preparation for East Link.

    1. The sign at UW looked different yesterday but I assumed it was just a faulty monitor. The main difference was no horizontal lines and a slightly different layout in the top section (the destination).

      1. Weird, I noticed there was a difference too. I didn’t know specifically what that difference was, but I definitely noticed something was altered. The Next Train times hadn’t been posted for a week or so, ST must have been preparing for this.

  9. This seems like a good time to eliminate the redundant and needless word Station from every train stop. Since trains only stop at stations (duh….), this word provides no useful information whatsoever.

  10. Does anyone know why the next 3 estimated arrival times are posted on these signs? It’s way overkill — too much useless information on signs makes them less readable and less effective. Moreover, integrating a future line’s arrival information will be that much more difficult due to the unimportant information already on the sign.

    If anyone can convince me that the next 3 arrivals are important I’ll stop complaining.

    As an example, here’s a link to an image that shows a sign from another transit agency:

    1. You can see the actual frequency of trains and their destinations (if there are branches or short trips).

      You know how long the wait is if you skip the next train for the one after to, say, get coffee, or if getting from the street to the platform takes time (like UW station).

      If the 2nd or 3rd train after the 1st train is delayed, you will know.

      The example you linked shows the next three trains on a single line. Train 1 and 2 are short trips and train 3 runs to the end of the line. In London, the next train (1) is always displayed while the sign cycles through subsequent trains or other messages on the other rows.

      1. I realized the example I posted was a single line that terminated at different stations. That said, I could find other examples showing only the next train on multiple lines that serve that station.

        The reasons you give for the additional information, spot delays or understand frequency, don’t resonate with transit riders, only transit wonks like you and I. Transit riders want to know when the next train is coming, and very little more.

      2. Transit riders want to know if trains are delayed.

        Transit riders want to know if they’ll make it to the platform in time for the next train and how long the wait is if they miss it.

        Transit riders want to know when the next train is if the one coming is too crowded.

        Not every system runs every few minutes such that showing only the next train is sufficient. Even the Paris Metro, with single digit frequency shows the next two trains. You can find some that only show the next train but the point is showing more than one is hardly “useless” and is very common.

        None of that is “wonky”.

        “understand frequency, don’t resonate with transit riders”

        Are you kidding me? Frequency is a big deal to transit riders and that’s one of the clearest ways to show it.

    2. It’s common on subway signs. Not necessarily in this horizontal format; some put each run on a separate line but the following one, two, or sometimes even three runs are visible.

  11. So will they write the line color in the future or not? Without colored signs, ST will have to add more text onto these crowded signs to define the color.

    It’s just one more reason why Link line numbers or letters should be primary, with colors as secondary. “L1” is shorter than “Red Line” is.

    1. They say the whole thing will be overhauled in 2023, and my guess is that means they’re going to replace all the LED signs with another fixture, either another kind of LED sign that can handle this better or LCD screens.

      It’ll be interesting to see what the Northgate Link stations get. U-Link stations got LCD signs but Angle Lake which opened afterward got the LED ones.

      1. I’m sure Northgate will get the same thing CHS and UW has. Those are full computers, which can display anything. Though Angle Lake got the old style signs even though it opened after UW.

        My guess is they will cram that into into the old signs. Maybe do something like “Airport (R) / Bellevue (B) : 3min B, 7min R, 11min B

    2. I’m hoping ST gives up on the horrible idea of a red “R”, a green “G”, and a blue “B”. Either put the whole color word or two letters or don’t put any word at all. Don’t make it look like these are lettered lines. Lettered lines aren’t based on the first letter of the color; they start at A, B, C, D. A blue “B” implies there’s an A line and maybe a C line. And those letters overlap with Metro’s RapidRide. The R, G, and B lines at Westlake will be a half-block away from Metro’s C, D, E, and H lines and probably others.

      1. In DC they abbreviate the color to two letters on signs and maps. This also aids people who are color blind.

        RD = red
        BL = blue
        GR = green
        YL = yellow
        OR = orange
        SV = silver

  12. now for real time schedule information for buses at stations; it is at UW, but not many others.

  13. Usually “Not Possible” from the tech department of a highly bureaucratic government agency doesn’t mean “not physically possible”, but “not possible with our current contractor arrangement”. Almost all hardware outside of maintaining the trains themselves is the purview of a contractor that has a monopoly on the product, and they won’t do anything without an additional contract, and they’ll set the terms (e.g. if you want added features on the signs then you need to pay us to buy new signs). You can replace the contractor when the contract is up, or pay a hefty fine to break your current contract, but then any new contractor will behave in much the same way because they know governments are slow-moving and politically-driven.

    So yeah you can pay a smart high schooler to make the changes with a $10 raspberry pi, after a weekend of dev work he could then probably upgrade all the signs at a station in a day, and have the whole system done in a few weeks, and he’d be happy if you paid him $30/hr. But no government would let that happen because work all must be done by authorized contractors.

    Authorized contractors will demand new signs (which they’ll make a percentage on), demand 6-12 months of development and testing time for a team of 2 programmers and an UX expert, then demand that each sign require at least 5 employees to change at $100+/hr each (1 person to change sign, 1 safety person below him, 2 safety people keeping others away from person making change, 1 on-site compliance manager).

    1. That is an oversimplification of what it takes to put together a passenger information system (see Glenn’s TriMet link) but I would agree that agencies, especially those of ST’s size, should reduce their reliance on contractors. Instead, they should hire and develop in-house talent.

      Did you know that they basically did what you suggested for those OneBusAway screens at downtown RapidRide stops? The system was put together by an intern at SDOT (not me).

      1. It’s not an oversimplification for what it takes to create a *display* system using data feeds that are already available. Obviously, tracking vehicles is an involved affair requiring equipment and communications in vehicles and along the route, but producing a simple sign that presents that data is not difficult. Working with an existing display, you are limited by the resolution and update limitations of the display, but you can accomplish quite with even relatively crude gear when your goal is to just throw some text around. Sometimes it takes some creative programming if the display doesn’t support cursor movement out of the box (rare on anything recent, but common on some very old displays).

        The problem is entirely institutional. It’s why Oakland struggles to fill in potholes to the point where angry ninja vigilante pothole repair crews have sprung up to illegally fix them in the dead of night. It’s why bike lane projects cost hundreds of thousands of dollars. It’s not that the government *can’t* do simple things cheaply and quickly, it’s just that they can’t do it within their own established rules.

        Here in SF people created vigilante passenger information screens scraping NextBus data long before Muni put up their own versions in stations, and they still do in places. One of the more frustrating aspects was how BART and Muni couldn’t work together, so always had separate information screens for arrivals, but the vigilante screens had no such restrictions.

        At one point Muni put in new color displays for ETA announcements over the stations platforms. It took them *OVER TWO YEARS* to actually enable them, so the screens just sat wrapped in plastic hanging from the ceiling, gathering pigeon poop. (It turns out some of them were defective with big chunks of dead pixels, etc. but I’m pretty sure they were long past warranty by the time they were finally turned on.)

        The platform arrival displays at Amtrak stations here still fail to display ETAs for long-distance trains. e.g. if you’re at Emeryville it will give ETAs for Capitol Corridor and San Joaquins but not Coast Starlight or California Zephyr. If you’re at a shared BART station like at Coliseum or Richmond, you have to physically walk between platforms to see ETAs for their respective systems. And of course good luck with bus ETAs.

        Fortunately they don’t control my smartphone, so I can have an app display all the ETAs for a station concisely (albeit on a small screen), but it would be nice to be able to just glance at the screens around me, and even nicer for the people who don’t know which app to look for or have a dead phone. =P

        Sometimes the red tape gets cut through (particularly in emergency situations), but for anything non-trivial the problem becomes that absolutely necessary oversight (engineering safety concerns, etc.) got intertwined with the politically-influenced regulations so thoroughly that it’s near-impossible to separate them. (Like, it would be awesome if we could have vigilante passenger trains but that has a far bigger safety risk than filling potholes.)

        It’s cool that RapidRide actually let someone set up some screens.

  14. “We do see issues with individual displays come and go. These are computerized devices, and like personal computers and electronics, they may experience software, hardware or – as with what we saw in the Lloyd District – connection problems. Some issues may take longer to fix than others. For instance a screen at SW 6th and Main Street in downtown Portland was stuck on the first scheduled bus arrivals of the day for weeks. Several attempted repairs and reboots later, we determined the motherboard needed to be replaced.”

    In the event anyone wants to read a bit more about the work that goes into these things:

    1. Maybe that’s the problem with the next-bus signs at Campus Parkway. The eastbound one has been stick at 10:37am some day for four years now, and the westbound one has a warning dialog in front of the text for almost as long so you wait for it to scroll around the dialog to find out when the next 49 is coming. I complained to Metro but they said that’s SDOT’s responsibility, and I haven’t gotten around to reporting it to SDOT but then I shouldn’t have to. Isn’t it reasonable for them to check the signs once or twice a year to make sure they’re still working? Maybe they’ve been fixed recently; I’ve stopped using those stops as often, but the last time I checked a few months ago they were still broken.

Comments are closed.