Regular Link riders might notice that real-time arrival information did not, in fact, come back with the Northgate stations as promised in September. ST’s John Gallagher says that “While the system was accurate the majority of the time, when it was wrong, it was really wrong.” It would cost “a significant amount of money” to correct these problems.

On the other hand, I saw real time data on the signboards at Westlake on Sunday, so maybe they’re under-promising and engineers have hacked something together?

Regardless, the good news is that ST’s PIMS program is set to wrap up in Quarter 2 of 2022, so the next-generation system will be here soon enough that it’s not worth it to patch up the old one. The result of an RFP in 2018, in 2019 ST shared some early mockups (at right in the picture above) and projected it to be ready with East Link. It appears we’ll get these many months early.

82 Replies to “Real-time arrival out for the winter?”

  1. I disagree that it’s not worth patching a system because the next release “will be here soon enough.” ST has fumbled this repeatedly since the line opened over 10 years ago. What magical thinking allows you to believe this time will be different?

    1. Throwing a bunch of money and effort at a problem that will go away anyhow in a few months would be a total waste of scarce taxpayer dollars.

      Better to work on the new system and get it right this time.

      1. ST’s mindset, effort, and priority should be on getting the next iteration of real time arrival information correct. It should not be on fixing, patching, or papering over the flaws in the old system when that system is going offline in a few months anyhow.

        Let’s be real here.

      2. This should not be a priority. Not when a solution is around the corner. It would be nuts to solve a relatively small problem, when there are much bigger ones that still exist. The escalators (old and new) aren’t in great shape, and the trains are not frequent. Run the trains every six minutes all day long, and then work on the reader board.

    2. If ridership comes back slowly, a lot more people will notice whether ST manages to get the where-is-the-next/currently-boarding-train feature going than how imprecise the RTA is. In particular, being able to transition from “next” to “currently boarding” will avoid confusing a lot of riders, at least southbound at Northgate through ID/CS. Loudspeakers will also need to be on their A-game with “Now boarding 2 Line to Redmond on the west platform”.

      Which reminds me: Start numbering the platforms (i.e. the sides). People can easily lose sense of direction down in those crypts. Tourists especially so.

      1. If it isn’t clear, the numbering would be Platform A and Platform B. Let the discussions begin about which should be A, which should be B, or if we should be calling them something completely different. Keep in mind how a 2-Line train that starts by going south ends by going north, simultaneously starting by going east and ending by going west.

      2. Here’s an additional thought on avoiding cacophany between simultaneous messages: Feminine (higher-pitch) voice for Platform A announcements and Masculine (lower-pitch) voice for Platform B announcements.

      3. The signs should simply read “Trains to …” listing both the last station, and one important intermediate station. Repeat for each line the station handles.

        The problem is MEN…. or more precisely, people who overthink their wayfinding with their sense of cardinal directions.
        (Case in point, the Everett Amtrak station. The route doubles back on itself at that station, so northbound trains are actually going southbound while they are accessing the station (and vice-versa))

      4. Agreed. It should also be short enough that it doesn’t need to scroll. For example, “To Bellevue, Redmond Tech”, or “To SeaTac, Angle Lake”. These could also change as they go further east/south. Once it hits downtown Bellevue, it could mention the last station (Redmond Tech).

  2. Seeing the sign mark-up jogged an observation. I must apologize for thinking that Redmond Technology is named Redmond Technology Center. I prefer the dropping of “center”. I suspect most of the public will just call it Redmond Tech. I note that Redmond is unique (and out of compliance with the new naming guidance) with three stations having the city name.

    Also, the nearby Overlake Village Station name will confuse riders going to Metro’s Overlake Transit Center that is at the Redmond Technology Station. Is the transit center getting renamed?

    1. Yes, the Overlake Transit Center’s already been effectively renamed. For the last year or two, all the buses and schedules have been saying “Redmond Tech Center” instead.

    2. It was called Redmond Technology Center in planning, so you’re not imagining it. People are using a variety of names until the lines open. Not everybody knows or remembers the new official names, or they think the listener might not recongize them, etc. That will all change when people actually alight there or go past the stations or see them on the map in the train.

  3. Why can’t they at least display the scheduled times on the boards? Link tends to stick to its schedule pretty well and scheduled times would certainly be more useful than completely blank screens for most of the next year.

  4. Sometimes I think that the best way to alert riders about trains arriving soon is to have a slow flashing of lights. I always thought that feature was great at DC Metro platforms when trains approach. I’d really like some sort of flashing beacon visible a block or platform level away, because it’s really frustrating to be approaching Link (like at the top of the stairs in the DSTT) and not be able to hurry slightly to make the train.

    1. If you are at platform level (and in the tunnel) you can tell when a train is coming from the gush of wind. I agree though, the main thing people need is a sense of when the train is coming from up above. Most of our station entrances are ways from the platform — it is really helpful if you have an idea if you should hustle or not.

      They could also run the trains more often.

  5. The 1 Line comes every 8 or 10 minutes depending on time of day, and the station loudspearker announces the intervals. I’d actually prefer maintaining headway over delaying just to get back on the schedule, so we don’t unnecessarily get more crowded platforms or trains. At least during peak, and before and after large events.

    I’m glad ST has captured the time savings from shorter deboarding/boarding times due to the pandemic, and look forward to the whining when the 1 Line takes more than 53 minutes on the current segment once white-collar commuters start returning to the office. I’m all for allowing the schedule to be based on a shorter trip time off-peak and a longer trip-time during peak based on data. We do that with bus schedules. We can do that with the train schedule.

    1. Does that work with the close interval for reversing at end points? If run 123 gets delayed four minutes and that four minutes is carried through the next several trains, won’t it get the terminal of the delayed trains fouled up?

      It seems to that if 123 were four minutes late arriving, it would have to turn in six minutes (four at peaks) which is pretty unrealistic without a driver break change.

      The damage would then migrate into the other direction.

      I believe that’s why most subway systems wait for the next slot if a train is delayed more than 1/3 of its headway.

  6. It’s quite the headscratcher how providing such a seemingly straightforward passenger amenity like real-time arrival displays has consistently befuddled ST for over a decade now. ST knows where trains are at all times, it’s no secret where the stations are….; I guess we’re all just missing something. Seems like the type of problem that could be permanently solved by throwing a couple staffers at it and saying: “Just get it fixed, yo.”

    My favourite chapter was when ST claimed the old LED displays in the DSTT couldn’t handle that kind of info… and then a few weeks (months?) later they were showing the trains times, lol.

    1. Current real time arrival software available off the shelf in the USA mostly uses satellite GPS, which doesn’t work in tunnels. For some reason, the signal system used for dispatching can’t be adapted to also provide this calculation, or at least no current vendors in the USA have.

      1. Seriously? The traffic control system doesn’t transmit train passage data from each turnout and control point?

        OK, I get that it doesn’t know which exact 1/20th of a mile a given train occupies. But it knows that it left the last time point at X and that it takes Y.Z minutes to cover the intervening distance almost every time.

        That’s good enough.

      2. The standard system available in the USA only shows block occupancy. It doesn’t tell you where the train is or even which direction.

        The outputs would be pretty meaningless for trying to determine real time arrival.

        It would probably be possible to put something together using block occupancy and the schedule, but it would be pretty custom. After quite a lot of expense and time creating it, you’d probably wind up with nothing better or sooner than whatever solution they are getting.

      3. The train to the satellite gates at the airport has a clock that counts down from three minutes to arrival. It can’t be that complicated.

        I am not sure you need a countdown clock with three minute frequency, but travelers don’t know the frequency is three minutes and are anxious. The clock reassures them.

      4. Glenn, if by “the standard system in the US” you mean “CTC” or “TCS” the control points transmit train arrival and departure information. Now maybe Link doesn’t have that sophisticated a system, but the turnouts are all powered, except possibly some within the confines of the Maintenance Facilities. So SOME communication between the control points and central operations must be occurring. You don’t want to throw a turnout under an occupying train.

        Even back in the 1950’s when CTC was introduced on mainline railroads there was a “graph” of train occupancy which was constantly produced. It’s a trivial programming effort to extrapolate the progress of a train from the most recent control point to every station it will pass.

        Obviously, as a train progresses over its route the extrapolations to the distant stations would improve in real time, likely to the point that for most actual arrivals at a station the prediction would be well within one minute. Once in a while, of course, something would befall the train between the previous control point and entrance to the station, and the sign would be wrong. After, say, two minutes, it should then switch to “Delayed”.

      5. Oh, and as to the “direction” issue, except in very rare instances, block occupancy on Link determines train direction categorically. It’s not “Two Main Tracks”; it’s classic “Double Track”, uni-directionally signaled.

    2. It’s not as straightforward as it looks. And ST’s culture is unable to prioritize passenger-experience things like this and invest in them right the first time. The new audio announcements, “Trains are running every 8 minutes”, are a welcome stopgap. Because the switchover between 8 and 10 minutes isn’t exactly at 3pm or 6pm, it depends on where you are in the line.

      1. “The new audio announcements, “Trains are running every 8 minutes”, are a welcome stopgap. Because the switchover between”

        Not really. An audio message every 8 minutes that says “trains run every 8 minutes” is just clutter. Nobody should be forced to interrupt their conversations because the train is running every 8 minutes.

        ST needs to cut out their obsession with periodic audio announcements. By all means, make an announcement when the train is actually approaching. Except in emergencies, everything else – including how often the train runs – should be signs. If it’s about blind people not being able to read the signs, that’s what braille is for – you still don’t need audio clutter.

      2. A lot of people want to know how long they have to wait, and “every 8 minutes” tells them the worst-case scenario is 8 minutes and the average time is 4 minutes. Plus if they’ve seen or heard no train while going down to the platform or if the platform is not empty, they can rule out the worst-case scenario if trains aren’t late. The announcment should even be repeated every four minutes. The announcements that annoy me are the endless repetition of “Sound transit does not tolerate harrassment”, etc. “Trains are running every 8 minutes” is what they should be saying often.

      3. No, that’s what signs are for. Announcements are for things where everybody needs to stop what they’re doing and pay attention, right now. Not for conveying general information that’s always the same throughout the day. An approaching train meets the bar for an announcement, as everybody is there to catch a train and needs to know which direction it’s going. “Trains running every 8 minutes” does not. I also have very little confidence in Sound Transit’s remembering to adjust that announcement if there were a temporary service disruption preventing the train from actually running every 8 minutes.

        It’s functionally like a posted flyer, but worse, since you have to wait several minutes to hear it instead of just reading it, and it’s obnoxious each time it comes on.

      4. I agree with asdf2 on unnecessary announcements. It is like an overly chatty flight attendant on a plane blasting inane comments to a captive audience.

      5. Yes, signs are ideal, but the signs aren’t working, so the announcement is a necessary stopgap. Are you also bothered when BART says, “Train to Dublin/Pleasanton in 7 minutes”? Again, it’s the generic administrative announcements that bother me. “Trains are running every 8 minutes” tells you whether it’s an 8-minute or 10-minute period, which changes during the day, and changes at your station at a not-so-obvious time. I even heard the 8 minute announcement once at noon.

      6. Speaking of, can we please get rid of “Sound Transit does not tolerate harassment” and the mask announcements? They’re almost as bad as “TSA has limited the items…” announcements in airports.

      7. “Yes, signs are ideal, but the signs aren’t working, so the announcement is a necessary stopgap”

        Only if you way over-engineer this. A better solution is to just take pieces of paper and staple them to the walls. It’s not that complicated, and it works just fine at bus stops throughout the King County Metro system.

        Imagine if Metro were to take out the paper schedules at the bus stops and replace them with an electronic voice that, every 5-10 minutes would say “buses run every 15-30 minutes”. It would be not only extremely obnoxious, but also much less informative.

        It is over-engineering and the insistence on electronic communication for information that is better expressed on old-fashioned paper that is the root of all these problems.

      8. @asdf2,

        I’m currently bouncing back and forth on the NYC subway system. Just spent most of the day with my riding the rails with my mother-in-law on various errands.

        She can’t read or write in any language, because in her culture educating a woman was considered a waste of resources. So it is hard to understand how stapling a paper schedule to the wall would work for her. Because it wouldn’t.

        We need solutions that work for everyone, not just transit wonks who can read a schedule

      9. Are you also bothered when BART says, “Train to Dublin/Pleasanton in 7 minutes”?

        No, because in that case it is telling you information you can’t get any other way. Frequency information is available on the walls, on your phone, on pieces of paper. It isn’t dynamic. Of course it changed during the day, but with a schedule and a clock and you have that covered.

        She can’t read or write in any language …

        We need solutions that work for everyone, not just transit wonks who can read a schedule.

        What if she doesn’t understand English? What if the announcement is in 13 different languages, but she doesn’t understand any of them?

        Come on man, we aren’t talking about critical information. Most of the people who wait for a train don’t check the schedule, because they figure it is a subway, and the train will be here soon enough. Those that are worried about it still don’t care. They don’t want to know if the train is coming every eight minutes or every ten minutes, they only want to know when *the next train* is coming, which is what this blog post is about.

        Oh, and it really doesn’t take a transit nerd to read a schedule. Ordinary people have been reading them for a really long time.

      10. I would argue you don’t even need to be proficient in English to read a train schedule. Numerals are nearly universal across world languages. Just read the numbers, and you know when the train comes.

        I could almost certainly read a train schedule in Japanese the same way, in spite of knowing zero Japanese.

      11. The schedules on the wall are posted once and get out of date. The audio announcement confirms what it’s doing now.

      12. @asdf2,

        Yes, the numbers are nearly universal, but being able to read a schedule is a step too far.

        Yesterday my wife tried to teach my mother-in-law her zip code. I’m not sure she was successful. And that is just 5 stand-alone digits. And my mother-in-law has been here for over 20 years.

        Things that seam simple to well educated, literate, and technically oriented people can be totally beyond comprehension to someone with no education.

        But a voice saying “trains every 8 minutes”? Or “next train in 2 minutes”? That is understandable to anyone who speakers English.

      13. But a voice saying “trains every 8 minutes”? Or “next train in 2 minutes”?

        Those are two completely different things. Most people don’t care whether the train is running every ten minutes or eight. A lot of people do know it, and those that don’t can find out easily. You don’t need to read, you can just ask someone. I know this is shy Seattle, but it really isn’t that hard to ask someone, especially if you are one of the few who can’t read a schedule, and actually care.

        The second is completely different. It is of value to everyone. There is no resource at all for you to find out this information — because the RTA system doesn’t work.

      14. “Most people don’t care whether the train is running every ten minutes or eight.”

        They care how long they have to wait! Every additional minute can be boring or frustrating and make people doubt the transit system’s quality. This varies from person to person and how they feel that day, but passengers in aggregate say waiting is the thing they hate most, even more than longer travel time, and especially when they don’t know how long it will be. The announcement gives an upper bound and people can estimate what point in that they’re likely in. If trains ran every 5-6 minutes it wouldn’t matter: you can jolly chill for 2-3 minutes. But when trains are running every seven minutes or less, it does matter. The wall signs aren’t ubiquidous, don’t tell you what’s actually happening now, and are sometimes wrong for a year or longer before ST updates them. They were wrong from summer 2020 to summer 2021 for instance.

        Again, the best solution is displays that say “Angle Lake 3 min”, but that’s what we don’t have, so the audio annoucements are second best. And this is more important than interrupting a conversation for five seconds. The annoying announcements are those about regulatory compliance that make it feel like a police state. “Trains are running every 10 minutes”, doesn’t do that; it’s informative and welcome.

      15. And just seeing or hearing a time announcement makes the wait seem shorter and less boring. Again, this is what passengers in aggregate report in surveys. “N minutes to next train” is best, but “every N minutes” is second best.

      16. Or “N minutes since last train”, as a few Link count-up clocks show for drivers [1], or as St Petersburg and Moscow have in every station.

        [1] One is southbound at Mount Baker at the south end of the platform, and there are a couple others along the line.

      17. @rossb.

        Please don’t be so indifferent to the immigrant experience.

        If you can’t read or write in any language, it is deeply embarrassing. Particularly in a literate and technically competent city like Seattle.

        And if you speak imperfect English as a second language, then you aren’t going to approach a stranger to “ask” a question.

        It’s just the way it is. Our transit system needs to work for everyone. And a voice coming over the PA giving some clue as to the wait time is highly valuable to some. Just as valuable as ADA compliance is to others.

      18. Giving the train frequency is great if you are making someone walk a ways to get to the platform. Seatac has such a sign.

        If you are at the platform it’s the next train announcement that is probably most useful.

      19. @Lazarus

        There are hundreds of languages spoken throughout the world. Sorry, there is no way to satisfy “everyone’s” needs.

        The belief that you can’t ask questions in broken English is absurd. I’ve asked for transit help in France and Italy and I sure as hell don’t speak French and Italian in any sufficient manner. Yeah, it’s harder. That’s life.

        Not everything can be perfectly suited for every person in every situation.

      20. Please don’t be so indifferent to the immigrant experience.

        I’m not being indifferent. That is insulting. This blog has a policy about insults like that, so please stop. If this was The Stranger, I would call you a sanctimonious prick — but I won’t, because that is against the rules. Please try and treat me with the same level of respect.

        And if you speak imperfect English as a second language, then you aren’t going to approach a stranger to “ask” a question.

        If you don’t speak English, then what good is an announcement in English? It makes no sense. You seem to think that people who speak imperfect English wander around, bumping in to walls, only to be rescued by their relatives. That’s not how it works. People with poor English skills ask questions in America — I’m sure it is how your mother-in-law survived all these years. You mean to tell me no one has walked up to you, in a very thick accent, and asked you a question in very poor English? Haven’t you ever traveled to a foreign country, where you don’t know the language? You get by with the help of strangers. Your French (or Italian, or Spanish) is terrible, but people figure it out, and point you in the right direction (often literally). This is how people get around, the world over. You really have it backwards. Announcements are difficult to hear, and if you aren’t sure about a word, you have to wait until they announce it again. If you ask someone, you can take your time, and keep asking until you understand.

        And a voice coming over the PA giving some clue as to the wait time is highly valuable to some. Just as valuable as ADA compliance is to others.

        Oh come on. That is ridiculous. ADA compliance is about mobility. It is about hobbling down stairs, or taking an elevator. It is about getting on a subway, or not.

        You are talking about an announcement that in all likelihood, is meaningless. Most people don’t care how often the trains run if they are committed to using it. They only want to know when the next train will be here. Let’s say they have two messages, back to back:

        1) “The train runs every ten minutes”
        2) “The northbound train will be arriving in two minutes”

        I guarantee you that 99% of the people don’t care about the first message. They either know it already, or don’t care. It gives you no clue as to the important information, which is the second message. You seem to think that somehow knowing the schedule — at that particular moment — is valuable, even though someone has already committed to taking the train. It isn’t. I suppose it might be handy for the way back, but guess what? It can change. So to be really proper, they should have an announcement like so:

        “The train runs every ten minutes until 2:10 PM. Then it runs every eight minutes until 6:20. After 10:00 PM, it runs every fifteen minutes. ”

        Yeah, that will be easy for ESL person to understand. Yup.

      21. I am in Honolulu. Yesterday we went to the hike to the top of Diamond Head. The fee is $5/person.

        The ticket machines accept only credit cards. No cash and no Apple pay.

        A young Japanese couple had only cash and Apple pay, and were confused by the process. They spoke almost no English and we don’t speak Japanese.

        But they did have a translator on their phone so they could explain the issue to us in English. I bought them two entrance tickets with my credit card and they gave me $10, and I showed them the entrance to the trail.

        I will definitely download a translator app if I travel to a foreign country and imagine that is the future, including translating images.

        Getting back to my analogy about the shuttle at the airport to the satellite stations. The countdown clock in red digital numbers and letters says “Next train”, with a clock counting down from 3 minutes above the doors to enter and exit the train. Even if you are illiterate you intuitively know what the clock means. After all, what else could it mean.

        Once on the shuttle the arriving gates are announced in three languages: English, Japanese and Spanish, although there are visual guides.

        Reading schedules is hard. Half the time they are wrong. Public announcements are often lost in the din of a light rail station, even if you understand the language.
        You almost always miss the first few words, especially when some of the announcements are totally irrelevant to your trip.

        Every restaurant has learned to create a QR code for their menu. With $148 billion why can’t ST create a QR code for schedules that a translator app would translate for the user? If ST ran a small deli it would be bankrupt.

        What any rider needs is a clear indication of the line and when the next train will arrive. A QR code on the train would be a good idea too. Otherwise everyone crowds at the edge of the platform or begins to line up, which adds to the frustration.

        This really has little to do with immigration status. Most immigrants have ridden the transit so they have some understanding of the route, but not arrival time.

        A traveler from a foreign country who doesn’t speak English and who hasn’t ridden the transit is the base customer. They want to know the line (ideally numbered since numbers are universal) and when their next train arrives. That isn’t rocket science, although no one has ever accused ST of being rocket scientists.

        All ST needs is a QR code for schedules, a posted written schedule for the very few without a cell phone ( even though Link is not a complicated system with basically two lines), an indication of the line, and a digital countdown clock — probably in red — because everyone will intuitively know what the clock means.

        Imagine if ST was in charge of the departure and arrival signs at the airport that are constantly updated, accurately, for literally hundreds of flights whose arrival and departure times, and gates, is constantly changing. Talk about chaos.

        The truth is ST — whether it is train derailments, budget estimating, or operations, is an inept agency. We got around $50 billion worth of light rail and bus service for $148 billion.

      22. @rossb,

        I stand by everything I wrote. I encourage you to reread it.

        As far as what it is like to be a totally illiterate older woman who only speaks very poor English as a second language, I’m sorry, but I side with my mother-in-law in this one.

        For her it is a lived experience. For you? Not so much.

        I encourage everyone to be a little more considerate and a little more empathetic. It only helps.

  7. Here’s my low-tech method for tracking trains. Just have the driver text someone in a control room with his cell phone each time the train stops at a station. The guy in the control room then manually enters the train location into a computer. The stations all have cell service, so this should just work, even in underground stations. This is not complicated.

    1. Having a train (or bus) operator actively using a cellphone while operating a train is a violation of safety protocols. Such behavior is verboten.

      1. Well, you don’t do it while the train is moving. You do it when the train is stopped at a station.

        The text doesn’t even need to be the full station name, just a pre-arranged ,two character abbreviation.

      2. They have radios, don’t they? Or some other way to communicate with the base and security?

    2. Or one of the station security guards could write down the time the last train passed on a whiteboard. Even passengers could do it if there were such a board.

      1. Maybe someone should create a crowd sourced app. I don’t know if a cell phone can detect where you are in the tunnel, but it seems possible. If not, riders could touch a symbol for the station they just used, when boarding the train. It wouldn’t be exact, but within a minute.

  8. For the time being, I’m OK with using Google Maps transit layer (presumably based on scheduled times) as a guestimate. Second quarter 2022 isn’t *that* far away.

  9. RTA has never worked well in this region, ST, Metro, whoever. I remember the first rollout back in 2001. It has never gotten any better. I don’t know if the issue is the transponders on the busses/trains themselves or the infrastructure collating and presenting the data, but I would be willing to bet that nothing improves in 2022.

  10. This is probably a good time to point out that the American Public Transit Association trade show is going on right now in Orlando (yea, in person). If anyone knows anyone (vaccinated!) down there that can sneak in and report back about what is available in this vein, it’d be interesting. My employer last let me go to one in 2005, after which I’d have to pay for it all myself.

    I did pay my own way to Innotrans in 2016 (employer paid for the €60 entrance ticket) to see all the great stuff we can’t have in the USA yet.

  11. What I don’t understand is that OneBusAway displays what seems like pretty accurate link information. AND, it’s not just schedule data but I often see it say “2 minutes delayed, 1 minute early, etc.” Where exactly is that information coming from that OBA displays?

    And if OBA has it, why can’t ST do it? Or if not, just put the OBA output up on their stations monitors…

    1. OBA’s data comes from Sound Transit. Sound Transit actually hosts OBA for the Puget Sound region.

      The screens aren’t your standard televisions or computer monitors, they’re specialty displays that aren’t very flexible in their inputs. Or, at least they won’t be flexible until the new system is finished next year.

      1. So there is some coded interface to the screens that the manufacturer refuses to release to ST? That MUST be the root of the problem, because no matter WHAT that code, a decent programmer could produce an interface to it in a matter of days. Four digit numbers and a station index comprise the data that are being transmitted…… Two attributes.

        Not hard unless the format and station indicators are withheld by the vendor.

      2. Possibly three attributes: you might need direction. However a more normalized system would have a separate code for each track at each platform.

      3. So the data is already being created and used by OBA. This is the same data that referred to by ST here? “While the system was accurate the majority of the time, when it was wrong, it was really wrong.” It would cost “a significant amount of money” to correct these problems.

        If that’s the case, the OBA data would be a whole lot better than nothing for the next interim period.

        But then that’s totally different from what you’re saying, that the challenge and cost is not related to the data so much as the “challenge” of getting it up on their monitors? Something doesn’t seem to be adding up.

        all the comments in this thread about “Oh

      4. Ah, Proprietary Software.

        Donchya just Love It!

        I feel your pain. Fixing my car would a snap if I knew what the opcode was, at least.

        So who is this divining programmer, sage, soothsayer that will be our salvation?

      5. Jim, if the interface to the displays has to be hacked, of course it would take longer. That was my point in replying to Nathan. It doesn’t mean that the basic point I made to Glenn that the TCS data stream provides adequate time/location information for acceptably accurate extrapolation of next arrival information, even if a digital report of block request for station departure is not collected.

        If it is not collected on a system as recently built as Link, one has to ask “Why TF not?”, but that would presumably be added to the new system. In any case, there are enough turnouts in the main stem at crossovers that decent extrapolation can be done today, as shown by the quality of the OBA reports. The only “problem” must be the interface to the displays.

        Needless to say, ST should communicate the situation to every potential future purchaser of displays of other technologies from that vendor.

      6. Tom/Ken, all I’m saying is that ST is dealing with tech that was designed to display a very specific input, and produce a very specific output, and to do it very well. They said it would cost half a mil to fix, which is likely some combination of new screens, new networking, and new data conversions, and paying someone to do it.

        Clearly, they think that having very wrong data some of the time is worse than having usable data most of the time, and they think it’s not worth the visible effort of fixing the screens a few months before the next system comes online with new screens anyways.

        I agree it sucks that it’s been decided that the Northgate extension has broken RTA until PIMS 2.0 arrives next year, but I think it’s unfair to ST to assume the problem is as easy as changing the channel from 2 to 3 and wiggling the antenna a bit.

      7. All trains will be Line 1 until well after the upgrade is installed. Apparently the plan is not to run turnback trains, so destination is not an issue.

        A half million dollars for this trivial interface development is EXACTLY why costs have ballooned so terribly. A lot of “friends” are getting fat off the public goose.

  12. As a software engineer, it is absolutely shocking that they could end up with a system that is so awful and unadaptable. Somebody made very bad decisions or are ripping off ST by charging them large sums of money for garbage (maybe the escalator contractor?).
    The excuse of “it’s complicated” or “it can’t be fixed” is something I have heard for many years from lazy or incompetent engineers and they manage to convince others that is the case.
    It’s not that hard of a problem and should not be expensive to get a great, robust, and adaptable system.

    1. Sound Transit didn’t install the displays in the DSTT. That was done in the 1980s before Sound Transit existed and nobody knew when/whether eventual rail would be built.

    2. I agree. Personally, I would like a law requiring all future software to be open source. It is much less likely that you will get crap if everyone can look at it.

      There are also bound to be people willing to fix or improve things for free. It doesn’t mean that all of the development is free. You still need people overseeing changes, and likely people being paid to develop some of it. But in the long run, it is cheaper, and a lot more reliable if projects are produced in this manner.

      Keep in mind, these sorts of horrendous software failures are common with government contractors. The CGI Group (a Canadian company) originally budgeted $93.7 million for Current estimates are that it cost the government $2.1 billion. Oh, and of course, it didn’t work that well (initially) which sullied the reputation of the program itself.

      This isn’t the first time CGI Group has done something like that, either. They had similarly inflated projects in Canada. These companies are really good at getting contracts, but not good at producing high quality, affordable software.

    3. Several European countries require government documents (word processing, spreadsheets) to be open source, and to use open-source software to produce them. That was set up twenty years ago and could be a model for transit-information systems.

    4. Currently, most of the new software development being marketed to transit agencies right now is “Mobility as a Service” stuff that is supposed to aim at integrating Lift, Uber, Via, etc into transit planning and payment, and has little to do with improving existing systems.

      (There is also some effort at turning transit signage into “infotainment” systems, but that’s aimed more at producing advertising revenue rather than useful information as far as I can tell.)

      So, see if you can get your employer to produce something better. There’s probably a ready market, but unfortunately the signage companies don’t seem to be headed that direction. So far, what’s out there strikes me as being developed by people that never use transit, and this have no idea what would actually benefit the riders.

      But, it really would be nice to see what’s being offfered at the APTA show this year.

      1. Well, the info part would be good. Can’t imagine what kind of ‘tainment they think would be a good idea.

      2. I wish I knew more about what the requirements are for the transit agencies. It would be interesting to see if it’s something within reach of competing in that space.

      3. “Currently, most of the new software development being marketed to transit agencies right now is “Mobility as a Service” stuff that is supposed to aim at integrating Lift, Uber, Via, etc into transit planning and payment, and has little to do with improving existing systems.”

        I sincerely hope this is not true. More of this kind of worst model public/private partnership will only result in a further diminished public transit quality. Orienting critical software infrastructure around such a thing would be downright reprehensible

      4. How about The Mambo Craze. (Bus at 1:03 and 1:29-1:41. Gondola at 2:13-2:29.)

        I don’t want station display graphics that like smartphones, or route displays that look like cartoons and have too much bright white background.

Comments are closed.