« A CTA driver's view of bus bunching | Main | CTA rail safety questioned in RTA study »

CTA driver gives us an inside look at bus bunching

Our CTA driver tells us more about bus bunching. Read his thoughts from yesterday.

The only way to really prevent a bus bunch at this point is for the following buses to stay back, but then this seems to only agitate the public whom are trying to get somewhere....which automatically rolls it into your NEXT question......

The comment that drivers purposely SLOW down is condescending and a little insulting. Not to sound as though I'm attacking anyone myself... but the thing that many of the public fail to realize is that we are bound by schedule adherence. Simply put, we are required to maintain the schedules that we are given for our runs.

We must regulate our time between time points. This is especially a sensitive on street with a big interval, say 20 minutes. The public KNOWS when that bus is supposed to come. Some days are very busy and require quick pick ups and go, some are "slower than normal" and may result in a bus "dragging the street". It's either on time (PREFERABLY) or late, there is no early.

If it is early, it must wait out the time. Early is grounds for write-ups, and it's not fair that any one of us must put ourselves in jeopardy for discipline because the public fails to realize this.

If a bus is killing lights or driving slow, chances are its close to its next time point a little too early. To better illustrate, I attached a copy of one trip from two separate runs that I have done. (Download it here.)

You may look at them and say "wow that's a lot of time if its not busy"....imagine how we feel. None of us want to "drive slow on purpose." Again, we are bound by schedule adherence, we HAVE TO DO IT. And in my personal experience...a lot of the runs do indeed have TOO MUCH TIME, and when there's a bus every 5 minutes and the one ahead is late, equalizing time is NOT an easy thing. Now imagine when it's so busy and traffic beyond belief....cue in the bus bunching again. See how it all connects?

Comments

One way I have read (source unknown) to reduce bunching is to just say that buses leave from the start point every X minutes. That way, they can travel at the speed of traffic (good or bad) and the driver doesn't have to worry about being "early."

I was hoping for a lot more from these two articles. You basically took the stand that it isn't your fault, and offered little perspective on what the solution could be. Except, to blame the passengers for not boarding quick enough. If there are 3 busses all bunched together, is there a way to communicate to the third and/or least full bus that they should take over as the lead bus? Make one of the busses run express and have the passengers exit who are getting off in a few stops?

See my post in the previous article in re: the things that are in CTA's control (writing good schedules, SOPs for when buses arrive at the terminal late, etc.) versus things that aren't.

Even the things that are theoretically under CTA's control, there are often budget constraints affecting full implementation (e.g. having a whole room full of dozens of people constantly writing bus schedules).

I have worked in a transit system (not CTA) and Chris is correct, the cardinal rule in most systems is DON'T LEAVE MIDPOINTS EARLY.

While this makes sense to me to have this on buses with, say 15-30 minute headways, it is nonsensical on buses with 3-6 minute headways like many rush hour heavy routes.

The CTA seems to try route adjustments all the time, so why not, for heavy rush period only, say all buses leaving terminus btwn 7-9 am and 4-7 pm, just let those buses roll as they will, no midpoint checks, and see if it helps alleviate bunching.

It could start on routes that have the trackers where passengers & CTA supervisors can see where a bus really is. Make it a "6-month experiment" like the reroute of the Montrose bus & Pink Line originally were so if it needs to be yanked, it easily can, but hey at least you tried listening to the riders.

Bunching isn't the problem. It's the symptom.

The problem is gapping -- the long time between buses, with a bunch behind the gap because the first bus on the schedule has been slowed down. This is really what riders hate -- they wait a long time, and then they see two or three buses together. If they hadn't waited forever, they'd care much less about the bunch.

What closes that gap ahead of the bunched buses is for the two buses to work as a team, getting the emptier second bus in front to carry more of the load. The second bus will have fewer riders exiting, so it can skip stops with no waiting passengers. If it does stop, frequently, the two buses can both use the time -- the front one loading while the second lets riders off, and they both pull through the light.

Slowing down the second bus makes the problem WORSE, not better. When you slow down the second bus, rather than having it pass, the first bus picks up an even heavier load, slowing it down more, and expanding the gap between it and the previous bus.

As others have pointed out, there are things CTA can do about bunching (some of which are already being done, like the spotlight the new administration has shined on on-time terminal departures); there are things that passengers can do about bunching (getting a Chicago Card, having fares ready, getting off at the back).

And then there is one thing the CTA could do that would make the problem worse -- slowing down the second bus in the bunch. Somehow, this driver has settled on the one thing that will make the problem worse as his answer.

Please, for god's sake, do not slow down the later buses in the bunch. Don't treat the symptom. Close the gaps!

Thank you ryan, I was all set to write virtually the same.
An overloaded & late bus should never stop to pick up passengers unless it's also stopping to let off passengers when bunched. The late bus should pass by the waiting people & let the empty follower pick them up.
What's missing is a method of the bus drivers communicating with each other to tell them of their passenger load.

Some of the delays are because some bus drivers are timid & some are more aggressive in moving back into the traffic lane from the bus stop. That too can cause a bus to be late. I've been on buses where the driver will let more than a dozen cars pass & wait for a break to get into traffic. Most limit it to 3 or 4.
California had a short time experiment with a large merge arrow on the left rear of the bus & it required cars to let the bus in when activated. There was a huge fine for disobeying, but the experiment ended with a whimper & wasn't made permanent.

I've noticed that, too, UCc, about getting back into traffic. My dad tells me that when he was learning to drive in our fine city, he was told (& I've probably mentioned it here before, but what the hey) that only the 1st car at a stoplight, the one even with the bus, had the right to go when the light turned green. The other cars behind were supposed to wait for the bus to pull out. I'll be generous enough to say the 1st 2 cars, then the bus. Don't see it happening these days, however, especially with moving all the bus stops across the intersection. But my dad still does it on the rare times he drives down here.

As for the communication aspect, I have to agree. I've seen radios/phones on the buses but almost never seen them used. PACE drivers used to be able to contact other buses back when I rode them to school (Waukegan area). You could ask your driver to radio your connection, and if the buses were close enough, they would wait for you. Stunned that didn't happen here when I moved into the city after college--and I mean the basic communication, not just helping riders make connections. Yes, I know NOW it probably has something to do with the number of buses out at any given time, and the relative frequency most of them come at (10 min, etc), but I didn't realize it then.
Would it be possible to assign a frequency to each route or is that too simplistic?

Couldn't a software application monitor the CTA bus tracker information, tell when busses are getting bunched and automatically send a message to the lead bus to not let passengers on for the next few stops?

Could work, Lance, but how do you stop passengers at those stops from getting on when there are passengers who want to get off? I see lots of angry people being denied boarding, especially in the freezing cold winter.

Having a nearly full, late bus pass by waiting passengers may help solve the problem of the vehicles being out of perfect sync, but it has the consequence of not being very passenger friendly. Is the job of the CTA to move buses, or move passengers?

While this is one way of dealing with the problem, and sometimes might be the best way, it shouldn't be the default method.

Here's one better method: Let's call the late bus A, and it's followers B, C, D, etc. A field supervisor sees that bus A has fallen so far behind that B has caught-up even while staying on schedule. The field supervisor then suspends the schedule for B so B can continue even if it gets ahead of schedule.

Early on, A might catch up again with B, but A should *not* leap-frog B. B will eventually pull further ahead because even though it's picking-up all the people waiting for the late A, as well as some who came for the on-time B, it doesn't have as many alighting passengers to begin with.

A gap will form, and perhaps C may catch up with A. If this happens near the end of the route, A should dump all it's passengers onto C, and short-turn so A can get back into possition. If this happens too far from the end of the route, suspend the schedule for C, and allow it to pass A just as B did.

Eventually, the idea is to short-turn A to get it back into possition for the return trip. If the driver of A has a required break, short-turn A soon enough so that work rules and safety concerns can be met.

As for the early part of the return route that A will miss by being short-turned, the field supervisor can instruct B to leave the terminal early, but to fall back to it's regular schedule upon reaching the point where A will again be working the route ahead of them.

Will this work in every case? No. Of course not. But it will work in many cases. However, it's too complex for the drivers to coordinate themselves, so this is where the field supervisors can become the most important operations employees at the CTA. And with Bus Tracker, they now can actually do this far more effectively than when they had to rely only on what they could see, and guess at what else was happening on the route.

===
The CTA seems to try route adjustments all the time, so why not, for heavy rush period only, say all buses leaving terminus btwn 7-9 am and 4-7 pm, just let those buses roll as they will, no midpoint checks, and see if it helps alleviate bunching.
===

I think this would also be a good idea. Again, not in every case, but certainly in some situations. This would be especially a good idea for a route that's primarily serving comuters from downtown, and has relatively few boardings once it leaves downtown.

This probably is something one would NOT want to try inbound in the morning, but again, the methods of relieving bus bunching should be many, and it needs to be recognised that not all methods are appropraite for all situations.

Mornings on these same kind of routes might be better served by a stratigicly placed extra bus ready to fill a gap that opens. Put the bus in the right place, and it can be at the ready for multiple routes. It could also provide faster replacement in case of a breakdown.

There are numerous ideas out there that are better than making people waiting extended periods have to watch the bus that finally appears just zip right by them as if they didn't matter.

Also, if you get into the routine of turning crowded buses into express buses, you give people an incentive to pack into a crowded bus instead of waiting for the next one in the hopes that the crowded bus will be turned into an express right after they get on.

===
California had a short time experiment with a large merge arrow on the left rear of the bus & it required cars to let the bus in when activated. There was a huge fine for disobeying, but the experiment ended with a whimper & wasn't made permanent.
===

This is the law in Oregon.

They also are designing new streets to be more tranist friendly. Instead of building bus bays to take the buses out of traffic that they then have to rejoin, they're building peninsulas jutting out from the parking lane to meet the buses in the traffic lanes.

Of course they don't do this on the main auto routes. It's generally done in the funky, local, neighborhood shopping districts. (BTW, all new street construction in Oregon has to include specific planning for bicyles: Either bike lanes, or adaptation of a parallel street to be a bike route.)

Portland may not be Chicago, but they are leaders in inovation, and are in many ways a real-life labratory for transit planning. A tranist professional would know that, and would spend a little more time studying real life results than making Powerpoints that try to convince us that inovation is occuring when it's not.

I've stood outside busses where the driver doesn't let anyone on (e.g. the 151 on Michigan around Chicago). It's not a big deal if the busses truly are bunched and another bus on the same route is right behind. People are receptive to waiting especially if they can see that next bus is less crowded.

I think part of the problem is too many stops; many routes need stops to be spaced somewhat further apart -and my other pet peeve, bus drivers slowing and/or stopping at every stop when no one is boarding - we need to bring back flag stops.

Of course, there is the bigger problem of antiquated routes based on long gone streetcar routes - the bus has some degree of flexibility that a, say, streetcar, doesn't have, why not use that to improve speeds etc?

I don't really understand the concept of a bus 'schedule.' There's just too many uncontrollable variables on a bus route. I can see having a time to leave the garage and a time to return, but that's about it.

I find it very annoying when a bus that is mostly full sits at several subsequent stoplights and then drives 10mph because it is running ahead of schedule. Trying to keep a bus running on schedule, especially during rush hour, is nearly impossible. What's the point on any route that runs every 15 minutes or less?

If the CTA reads this please stop this policy, it confuses and angers riders. I've never heard a single driver even announce why they are crawling along their route. No one in their right mind uses bus schedules during rush hour anyways. I like to get places quickly as does everyone, so when I get on a bus that is running ahead of schedule I get punished and have to watch as any of the 3 other bus routes that I can take pass my ahead-of-schedule bus.

I think I am going to start asking drivers when I board if they are running ahead of schedule and if so get on the next bus that comes by.

Just a FYI, you can tell if they are running, late, early, on time, by taking a look at the monitor over the drivers head. It says Late;: x minutes, etc...take a look next time..

KevinB

KevinB: I have looked at it a few times after I saw you mention it in another post. I wouldn't be able to tell quickly before boarding whether it is early I don't think.

Cheryl: Neither do the drivers! Ha ha ha ha! (And more seriously, Huberman seems intent on doing away with schedules. So at least one person thinks that's the solution.)

Actually, I've got a lot of respect for drivers (moreso since I gave up the Red Line for the 147), and it's always the worst-case scenarios that we remember. I first learned about the policy of sitting at stops to stay on schedule on LiveJournal's CTA community; I've only ever seen it on the 80 Irving Park, and not once has a driver ever announced what's going on. It's just one of the things drivers *can* do to improve communication. Maybe most of them do.

Bus bunching will never go away. There are too many causes. It's like trying to cure the cold. Complaining about it is just, as I've said here in the past about every other topic, venting and whining. We feel a little better but we haven't contributed a thing.

Once your favorite routes to gripe about are on the Tracker, save the bookmarks on your cell phones and use 'em. For all the tiresome blather from some people about the Tracker not being a tool for riders, it's all we've got, and it stays up-to-date in real time. It works great for me.

Tracker wasn't working this morning for the 92. It said "No information available." Very helpful!

(Also, please keep in mind - these tools ARE useful (when they work) but there are TONS of people who either don't have a cell phone, or, like me, have one that can't access the web.)

I always thought the easiest way to deal with bunching was to simply "express" the first bus(the crowded one). Allow the first bus in line to make no pickups, only drop offs until the bus is either safe to put more passengers in to, or significantly out of the bunch.

Far too often at belmont and sheffield you'll see people cram the bus so full that the last person on board cant even get to the farecard box. This is not being safety concious and it wont take too long for the CTA to get into trouble with this kind of thing.

Rusty, thanks for a better restatement of the issues. After posting, I realized that my terms 'first' and 'second' bus were problematic, and A & B would have been better.

DavidJ, how do you tell people who are waiting for the bus that they can't get on, it's an express, they're only dropping people off? Because I know I'd be PO'd if a bus stopped while I was waiting for the bus and I couldn't get on it.

Cheryl, for the sake of running the buses on schedule, people would be a bit more receptive to the idea, especially if they're bunched.

The buses could have an announcement that says, "This bus is running express due to the close proximity of the next bus. Passengers will be allowed to disembark only. Please be patient as the next bus will be here shortly."

@ nd:
I completely agree that stops too close together. Every intersection does not need a bus stop. I site the example of the Belgrade, Serbia, bus system. Bus stops are a good 300-400 meters apart.

J, a terminal timer pilot was implemented on the 72 several months ago. It helped a little bit, but bunching/big gaps (since they are pretty much bedfellows) is usually exacerbated as the route progresses. Terminal spacing only temporarily solves the problem.

Moby,

It isn't quite right to say that terminal spacing only temporarily solves the problem. It does help, it always help, to have the buses leaving spaced properly. If they aren't spaced properly, bunching is inevitable.

However, there are other causes of bunching, and on-time terminal departures can't help this.

The other thing is that there's a difference between leaving on time and spacing. If bus B was supposed to leave 10 minutes behind bus A, but left 17 minutes behind, it does absolutely no good to "space" bus C 10 minutes later. All this does is insure bus B will be disastrously late before C can get in a "bunched" position where it can help.

Again, the problem isn't the bunch. The problem is that 17 minute gap. A 17-minute gap will almost certainly become a 30 minute gap before the run is over, unless C can pitch in and take some of the load. You have to attack the gaps, not the bunches.

In short, gaps cause bunches, which look awful. But bunched buses, if they work together, help keep the gap from getting worse. People hate seeing bunches. But mostly they hate being late, and it's the gap that makes them late.

So depending on what the 72 "terminal timer" pilot did, it may have made service much worse.

You don't need a starter if the drivers have a copy of the schedule, and follow it. A starter is nothing but a babysitter on a route with a schedule.

If drivers aren't starting on time, a field supervisor needs to find out why. Is it an attitude problem? Or are the schedules so bad that buses aren't reaching the terminal early enough for any required breaks to take place before the scheduled departure time comes.

If there is some cultural issue that's at the root of late departures from the terminal, a starter babysitting the drivers is one way of dealing with the issue, but that can't be the only management tool being used, either. At best, a starter is a short-term way of dealing with driver performance issues, but a starter doesn't help bus bunching, especially if the schedule the starter is working from isn't realistic to begin with.

ryan,

What I meant by "temporarily" is this: it only further delays an inevitable tendency for that route to bunch as more dynamics are inserted into the system.

I think you're attempting to simplify bunching when it's really a chicken-egg dilemma: gaps cause bunching which causes gaps...it doesn't matter which event causes which; both must be attacked at the same time.

what noone is really understanding what was stated in this comment from the driver noone is to blame it is something that happens. buses cannot weave in and out of traffic they are meant to stay in the right lane most of the time so they can pick up passengers . now think about rush hour in chicago or anywhere else most people in the right lane are going slow cause of traffic ahead and others turning. so if you left your house ten minutes before another car but picked up more people along the way wouldnt that car behind you eventually keep up. now allso say you had to reach a certain block within that route at a certain time five minutes more or less and you'll get a ticket too many tickets no more driving so tell me whats the answer to that.

how to be a cta driver.

Post a comment

Comments are moderated, and will not appear on this weblog until the author has approved them.

If you have a TypeKey or TypePad account, please Sign In.

TrackBack

TrackBack URL for this entry:
http://www.typepad.com/services/trackback/6a00d83451c39e69e200e5537ac9198833

Listed below are links to weblogs that reference CTA driver gives us an inside look at bus bunching:

Share news tips

Elsewhere