CTA Tattler

FYI, she calls them back holes, not black holes

Carole Brown now channels Orwell & Newspeak!
Prediction accuracy!
That's a sick joke!
What it is, is that the 22 & 36 buses are so late at times that the CTA is embarrassed to let the public see the actual results.

It also means, as I've been saying for the last couple of weeks that bus tracker is alive on both routes, but not for the public, only the Orwellian weasels at 567 Lake St.

Yesterday, Sunday, Nov. 2 at 3:15 PM there were three 22 buses NB at Pratt & Clark within a minute after at least a 20 minute wait. It might have been a longer wait, but that's how long I was there.

Any interesting bus re-routes for anyone today? My 136 exited and North/Lasalle instead of getting off at Wacker. The bus driver said that route (or Wacker) was closed already for the Obama rally. Anyone else experience this?

It looks like your bus operator jumped the gun. According to the CTA's Bus Alerts
http://www.yourcta.com/news/whatsnewA.wu?action=displaynewspostingdetail&articleid=121883
posted yesterday, the reroute shouldn't start until 10 am today.

Awwwww, poor baby had to wait 20 minutes for the 22 bus. Give me a break!

I, for one, continue to think it's a *good* thing that the CTA isn't knowingly putting bad information on the Tracker.

And "prediction accuracy" sounds like a straightforward term to me. It's not designed to mean anything other than what it says: the data they get back from 22 test runs is unreliable.

I see no use in accusing the CTA of trying to cover this up when they're *admitting* it doesn't work right. That's sort of the opposite of a cover up, when you think about it for a second or two.

That makes sense. I didn't think the streets would be closed off so soon. You're right, he jumped the gun by a couple hours.

Dave:
1. Three buses in a row means about a 30 minute wait, the headway on Clark is about 10 minutes.
2. I wasn't waiting for the bus, I was sitting in McDonald's there looking out the window. I knew exactly when I walked in there, it's on the receipt & I looked at my cellphone to check the time when I saw two buses right behind one another & a third just a block behind those two.
3. Get off your high horse, I don't expect the perfection of Hong Kong or Singapore, just a competently run system, which the CTA is not!
4. Prediction accuracy doesn't mean the data is unreliable, it's a cover your ass term meaning that the data isn't what you want so now you have to "massage" the data to get it to where you wanted it.
The Pentagon does this all the time with its testing procedures, where failures are turned into successes.

Either the buses are where bus tracker shows them to be or they aren't.
If they are, then bus tracker works!
If not, then Brown should have said so in plain English & not couched it in gobbledygook jargon!

[Either the buses are where bus tracker shows them to be or they aren't.
If they are, then bus tracker works!]

Except that you are familiar, I assume, with the part of the tracker that PREDICTS how long the wait is until the next bus arrives. If this data is not ACCURATE, you have a problem with "prediction accuracy."

It's not as simple as knowing the location of the bus. And there's no indication from Brown's post that there's any "massaging" of the data being done, except for the obviously justifiable step of adjusting the schedule.

There's no reason the CTA couldn't post the locations of the busses on the bus map and simply avoid making predictions as to arrival times. The algorithm may be confused by routes that start and stop at the same place (though that sounds awfully weak to me), but I am not -- tell me where the bus is and I'll have some idea when it will arrive.

I once had five consecutive trips on the 36 end at Foster with no notice, stranding everyone up to a mile and a half from his or her stop. It's not entirely implausible that the CTA would prefer not to publish documentation of that kind of thing.

There's no reason the CTA couldn't post the locations of the buses on the bus map and simply avoid making predictions as to arrival times. The algorithm may be confused by routes that start and stop at the same place (though that sounds awfully weak to me), but I am not -- tell me where the bus is and I'll have some idea when it will arrive.

I once had five consecutive trips on the 36 end at Foster with no notice, stranding everyone up to a mile and a half from his or her stop. It's not entirely implausible that the CTA would prefer not to publish documentation of that kind of thing.

UC,

You always sound like a turbine about to blow a bearing. Can't you hear yourself whine? Like always, if you can do better, step up. It gets old to hear you always complain about everything. Does a day not go by that you don't have the CTA rain on your personal pity parade?

This blog wouldn't exist if the CTA was any good!
So of course I complain about what doesn't work with the CTA, which is just about everything.
And as Eli wrote, prediction isn't necessary.
Just show where the bus actually is!
That's all I want or need.

And we never heard about prediction accuracy before on any of the other routes with bus tracking, just the 22 & 36.

Now why is that?
Maybe because they can't figure out how to explain the inconsistencies in the operations of both routes.
1. If bus tracker shows a NB bus at Foster & Clark at 10:32 AM & it's there, then it works.
2. But if it shows that same bus at the same location, but it's actually at Belmont & Clark, then it doesn't work.
3. And last, if it shows that same bus at Foster & Clark [along with three others] at 10:32 AM, but the schedule has it being at Touhy & Clark, then it's also working, but embarrassing to the CTA.

I'm guessing #3 is the real reason bus tracker for the 22 hasn't gone public!

I thought the same thing when I read it.

prediction accuracy? lol

I still want some of what the folks at the CTA are smokin'

KevinB

I'd like to know what YOU guys are smokin'.

Bus tracker can't tell you how long the wait for the next bus is by tracking where it is and how fast it's going. It can only do so by tracking where it is and when it's 'supposed' to reach a certain bus stop - i.e. a pre-determined schedule. If they're looking at the printed schedule and seeing that they're useless 'cause the buses are always bunched, then of course it's not gonna work until they figure out a realistic schedule, and what's the use of adding a route to bus tracker if it'll only give you false information?

I hate the waits for those two buses, too, you guys, but to suppose that they're doing this on purpose (or, for that matter, without good reason) is just piss in the wind.

Do you guys really think, with everything going on with the CTA right now, that whether a single bus line runs late or shows up at the right time in the right place is going to cause them some overwhelming embarrassment? You guys need to take a step back and think about what you're saying, or at least implying.

AlsoDave: Not only does it cause embarrassment at the CTA, it gets them in hot water with King Richard & his Quixotic quest for the 2016 Olympics.
Everyone knowledgeable in the Olympic world has said that Chicago would have it sewn up except for our wretched public transit system.
For all we know, the Lords of the Rings are also viewing bus tracker from Lausanne!

"This blog wouldn't exist if the CTA was any good!"

Actually no UC it would. Kevin stated purpose for this blog is: En route, we see the strangest things and hear some very interesting conversations -- many of them one-sides, thanks to the proliferation of cell phones. This blog will document things seen and heard on the CTA.

That last sentence is the most important. Kevin created this to just document crazy commuting tales. It was up to him to change the scope to more about how the CTA is run. But I am pretty sure that he didn't create it to for a place for your personal complaints and whining about the CTA. If I have that wrong please correct it Kevin.

And you could at least try not to sound so angry and crazy with your posts. KevinB complains just as much but at least most of the time he can be polite. At this point I am basically hoping for the worst commute possible for you and I am sure I am not the only one. If the CTA is so horrible what have you done besides complain here about it? Have you tried going to the meetings, contacting the press, anything besides complain?

One comment on "prediction accuracy". Ask Jeanne Dixon about that one. She was right one time ( Please, Mr. President, don't go to Dallas. ) and made an entire career from it. Granted if you're going to be right make it signifigant and memorable.

There seems to be a belief here that the bus tracker system *knows* where each bus is at any given second. But the fact is that the GPS isn't sending data multiple times a second, nor does the data it sends always make it back to the servers.

At any given time, many buses on the maps are shown where they are using a calculation based on where they were the last few times the data was successfully collected.

Thus "prediction accuracy" isn't just a matter of providing the estimated time of arrival at a future stop. It's necessary to place the bus on the map accurately in the present as well.

What you would see if they don't get this "prediction accuracy" problem fixed is buses jumping back and forth on the map. It can sometimes happen now, but since the goal is for it to happen so seldom that you don't notice it, well, you don't notice it.

When they invent a GPS signal that never has a reception problem, then prediction accuracy won't matter.

Anyone here have satellite radio? Does it ever cut-out? That's the kind of thing I'm talking about here, except that your satellite radio can have a few seconds of programming buffered in anticipation of losing the signal as the antenea moves. Prediction accuracy is the abiity to predict accurately where the moving antenea on the bus is when the signal isn't there.

Thanks Rusty that was helpful and will hopefully stop the conspiracy theories of UC.

(apologies for the double post above)

There's no reason the CTA couldn't provide -- either on the map or as a feed elsewhere -- each location-timestamp pair as it comes in from the GPS. The fact that the data are necessarily imperfect and have gaps is not a reason to withhold it from interested parties. It's nice to know where the bus is right now, but if all that's available is its location as of five minutes ago, that's much better than nothing.

Are these GPS signal problems something that's happening on all the other routes? First I've heard of it...first anyones heard it.....

Again, if this was a circulator route with its own problematic predictive issues, I could understand.

Also, by now they should have this down to a science. It's not like it requires a rocket scientist to install and maintain these devices. From information we've got from Ron before, this is just adding a modem to transmit the information already available in the in-bus bus information system.

As it's been stated much more eloquently that I ever could, if if the "predictive" part of this is not perfect, there is still much value in the bus number, location and destination of an approaching bus is of value even if there are some bugs on the predictive part. There are still ghost buses on the routes that have been in operation for a while....Destination information (is this bus going to the howard red line station or just to foster?) is still very salient.

As one of the resident cynics, I find it highly unlikely that these routes have unique problems as related to the bus tracker, I do however think they could be highly embarrassing Olympic wise and would shed a lot more light on just how sporadic and unreliable these routes are....

KevinB

Reality Check: I highly doubt that the BusTracker being implemented on 2 North Side bus routes that will most likely go nowhere near any event sites will have an impact on the Olympic Selection Committee. The only potential Olympic site served by either of these routes serve is Wrigley Field. And maybe not even that since baseball is supposed to be eliminated as a sport for next time.

Correct me if I'm wrong, but the GPS systems are built into the buses, as I understood that was how they knew where the stops were and called them out. The only part that wasn't there was the data transmission device and it's interface to the in-bus informations system.

Depending on the transmission method used, the the reliability of getting data from point A to point B is very high these days...

KevinB

KevinB,

Here's how I *think* this works: The GPS systems used for the stop announcements are receivers. They receive the coordinates from the satellite, and the software matches that with the recorded announcement of the upcoming stop. For bus tracker, they need some kind of transmitter on the bus to send the GPS data back to the CTA server. It's kind of like the difference between pagers and cell phones.

Has it occurred to anyone that the prediction accuracy for the #22 and #36 are so bad because the traffic on Clark and Broadway is so horrific? Hopefully the CTA won't decide that since the traffic is so bad, they have no business running busses on those streets, and they cut the routes entirely.

The CTA is exploring adding wings to the 22 and 36 so that they can bypass UCC.

"It's nice to know where the bus is right now, but if all that's available is its location as of five minutes ago, that's much better than nothing."

Eli and KevinB, I would have to disagree. Then the CTA would have to say, "you can depend on all our bus tracking, except 2 routes." That would just confuse people especially people who don't ride regular. Plus if they cant get an accurate prediction who's to say that the bus really was there 5 mins ago. It could of been 10 mins ago or long if it really is having trouble with getting the signal out/traffic. It would be a horrible PR move for the CTA to put a half route on the bus tracking system.

a rider: I hate to nitpick, but that's not how GPS works. "The satellite" has no idea where anything is. The GPS receiver depends on getting signals from three or more satellites in order to calculate its position. When driving in the city myself, my GPS receiver will lose signal quite often or get confused as to which street I'm actually on. It all depends on terrain, weather, and nearby architecture.

Algorithms, shmalgoritms. I aver that the probablility of the 22 and 36 getting on Bus Tracker is directly correlated to the probability of the Cubs winning a World Series.

I have to agree with Lone Gawker.... If I was waiting for a bus with UC I would gladly wait 45-60 minutes just to see him get all mad. Man that would be great....

Lone Gawker, really what I'd like to see isn't two routes that work differently on the regular bus tracker interface, but access to the underlying data for all routes. Third parties could do interesting things with them, including things that would be useful to those of us who ride the 36 and 22. Right now the CTA is saying that we can only have the finished product, and if the data for a route are insufficient to create that product then we get nothing. There's no reason it should be locked up in the vendor's proprietary systems.

Eli
What would individuals be able to do with the information? Most people have not touch any kind of serious math since high school so how will they be able to make a truly predictive tracker better than a company that actually employees individuals to do that? And also we don't know if these routes have some serious signal issues because of the buildings/structures they encounter on the route. I rather know that the products works than have some algorithms on the CTA about the route.

Oh and I would still like to hear how UC is working on changing the CTA instead of complaining about it.

"If this data is not ACCURATE, you have a problem with "prediction accuracy.

"It's not as simple as knowing the location of the bus."

Well, if I know that the bus was last seen 6 blocks away 2 minutes ago, *I* can predict when it's going to get to me, even if the bus tracker can't. So the data would still be useful. Still, I can understand why the CTA wouldn't want to put the route on the system if a major component of the BT system doesn't work right with respect to that route.

Some of the posts above suggest that the predictions are just a function of bus location data. I thought the whole point of connecting the gadget to the bus in some apparently labor-intensive way was so that it could also be attached to the bus's transmission in some way and collect information about how fast the bus was traveling, how much stop-and-go there is, etc. No?

Chris: It's not that the 22 & 36 will serve the Olympic venues, it's the general decrepitness & incompetence of the CTA that the IOC has been looking at, commenting on & worrying about.

My sole reason for wanting the Olympics here is to force Daley to get the CTA running the way it did 20 years ago, not anywhere near perfect, but it was acceptable, except for the lack of air conditioning on the buses.

Rusty: I don't care if the location were to jump around a bit at times. I don't even care if the locations shown are up to a quarter mile off, which would mean the bus was in an area up to a half-mile long.
I will be happy with a general indication of where the buses are.
Put a notation on the site that there are problems with this particular route showing up correctly.

Otherwise, I am going to stick with the belief that the CTA doesn't want the riders to know how messed up the 22 & 36 really are.

If the 157, which has almost all of its run through downtown high rises can have bus tracker successfully, then any route properly equipped should be able to.

And I will never drink Carole Brown's Kool-Aid.
She drives to work every day - in the Loop!

>>>
Are these GPS signal problems something that's happening on all the other routes? First I've heard of it...first anyones heard it.....
<<<

Nope. It was part of the original description of how it all works. You've just made a lot of assumptions that aren't based in fact.

This is also why they never considered retro-fitting the oldest buses. It wasn't just a matter of securing a device to a flat surface, and installing an antenea. The device has to integrate wtih other systems on the bus to determine things like actual ground speed, average ground speed, and ranges of ground speeds. And that's all to tie into the predictive nature of the system.

They never said this was pure GPS. They've always said that it's a combination of GPS and other predictive factors. No one paid attention, and I was treated like I was an idiot when I tried to explain this months ago, too.

I'm not making any of this up. This is all extropolated from official CTA communications -- parts most others ignored or misinterpreted, but none the less were there.

This is NOT the first time you or anyone else has been told about this.

To everyone who thinks it would be of some use to know where the bus was 5 minutes ago, I'd have to say you are nuts. Inaccurate information is worse than no information at all. You may think you want to know these things, but if that's what they actually gave you, it wouldn't take you long to see just how useless it would be.

"My sole reason for wanting the Olympics here is to force Daley to get the CTA running the way it did 20 years ago, not anywhere near perfect, but it was acceptable, except for the lack of air conditioning on the buses."

HA, hahahahaha, The CTA was WAY worse 20 years ago, just more people live in dense north side areas take it now. The a/c didn't work, yes, but then almost every bus had windows flopping around, water leaks and was generally about to collapse.

Does anybody know if the 22 and 36 have more buses which are about to be retired?

I notice there haven't been any more suspension complaints about the LSD buses - I wonder if people realized that it was in fact, indeed... the paving, not the bus itself....

"I will be happy with a general indication of where the buses are."

Will you though, UCC?
Will you really?

Yes I will, really!

And there are only New Flyer buses running on 22, 36, 151 [the North Park runs] & 155.
I haven't seen a TMC out of NP in a couple of months, but saw a couple in the Loop on some other route two weeks ago..

And I forgot, it was the L that ran much better 20 years ago, not all of these slow zones. I don't remember that much about the buses except for the lack of A/C & the windows. I avoided the bus most of the summer if possible.

[As it's been stated much more eloquently that I ever could, if if the "predictive" part of this is not perfect, there is still much value in the bus number, location and destination of an approaching bus is of value even if there are some bugs on the predictive part.]

I can understand this reasoning, at the same time, from a customer service standpoint I think it's a bad idea to have the Bus Tracker offer different functionality for different routes.

You might have the patience to sift through "imperfect" data, but I don't - I just want it to tell me when the next bus is supposed to be there. And while this is obviously speculation on my part, I assume a large majority of the public is the same way I am; they just want it to work, dammit.

And again, just on principle, I feel that offering information known to be "imperfect" - which, of course, is a nice way of saying "wrong" - is worse than not offering that information at all. Doing so is more likely to add to customer confusion than it is to clear it up.

UCC, when you say,

"I am going to stick with the belief that the CTA doesn't want the riders to know how messed up the 22 & 36 really are."

Maybe it's just the wording, but it comes across almost as if you think you're the only one with this knowledge. Wouldn't any regular riders be privy to the frequency of buses on those routes? And aren't there enough regular riders that those routes' gapping issues are therefore pretty much common knowledge? What riders lack this knowledge?

I'm pretty sure that it was explained previously that the way that the bus "knows" what stop it's coming up on is using GPS data (from 3 or 5 or x sats, don't know, don't care, it still knows what the next stop is)...the explanation from Ron was that the part that wasn't there was the interface to the already present onboard system and the radio modem transmitting the raw data to the central server and that the "predictive" part was handled by that server.

Again, I can't imagine what "predictive" problems there could be on a straight up and down North/South route, not a circulator, not anything different that any one of the number of other routes that are already available on Bus Tracker.

I already understand the problem of "ghost buses" when they are not logged into the onboard system that calls out the stops and activates the overhead display.

Again, me knowing that bus #xxxx, going to Howard Red Line station (as opposed to turning off at Foster) and whatever "prediction" it has toward the possible arrival time for the next several buses is of value to me (and I'm assuming others) from the discussion.

So, my bull**** detector is going off full blast.

What makes the 22 Clark and 36 Broadway so much different than say a 77 Belmont or a 152 Addison or a 151 Sheridan? Inquiring minds want to know.

KevinB

What makes them different? My guesses, and the order in which they likely matter: More dead spots. Less predictable congestion. More stops actually made.

What's made them so different all these years without the bustracker?

Does anyone else on here really feel like the 22 and 36 are that horrible? It seems like the only ones that complain about it are KevinB and UC...

And again imperfect info is not better than no info. People will complain when the tracker says a bus is only 2 blocks away when in reality it passed you up 5 mins ago. Who really wants to sit and try to figure out how accurate the tracker is from day to day. Not any sane riders.

Bob S.:
I don't claim any special knowledge, but long years of observing government have taught me they never release any info that doesn't meet their expectations or will cause public ridicule or embarrassment.

We all know when one or two buses are at a specific location.
But I can't & I doubt if there is anyone else that can log when every bus on a route passes a specific location.
And none of us can log where & when every bus on the route is except the CTA. It alone has access to the raw data.

Even if the bus tracker for the 22 supplied to us was "off", if I could look at the entire route & see that all the buses are spaced out relatively evenly, I would know that it's running about on time. But if there was a large gap & several buses were bunched up, I would be able to figure out that something is wrong.
I might not know exactly when bus is coming, but at least I would know that I wouldn't have an unreasonable wait.

If we waited for perfect data, we wouldn't do anything.
Ever!
No one ever has perfect data & no one ever will!

Bus Tracker successfully works on 75 routes & it won't on just these two?
HAH!

And Lone Gawker: Don't set up a straw man, I'm not going to waste my time with that foolishness.

Thanks, UCC. To be fair, I don't think many institutions, companies, or people give out embarrassing information about themselves. And my impression is that if they showed you what you're suggesting, that display could still be significantly off. Keep in mind that people here and elsewhere have reacted with surprising vehemence when the tracker is mildly inaccurate, which happens routinely for a number of perfectly reasonable causes. GPS signals do drop out in high-rise canyons; buses do miss lights; wheelchair users do take three or four minutes to get on and secured, longer if the bus is crowded. And even its guesses are inaccurate enough that I've wished it showed me not only the estimated arrival time but the stop the bus is currently at, so I could judge how far off that display is. My experiences at the beginning of routes are that that's where the tracker is most inaccurate; buses show up on the display without being active on the route, and I've seen arrival times for those buses be anywhere from 6 to 25 minutes, but it's never accurate until the bus leaves its terminus.

I'm not arguing against any of your points here, I'm not suggesting you don't know all this, and I'm not suggesting people who use the tracker for other routes don't have experiences that contradict mine. It's just one perspective that the tracker is a mildly accurate and often useful but always flawed tool.