Tackling Train Crowding from the Inside

Crowded CTA Train

Crowded CTA Train. Credit: flickrhivemind.net

This sight is not uncommon during rush hour on any urban transportation system. I’ve traveled on many public transportation systems in the world, but the Chicago L exhibits a pretty unique phenomenon during periods of high demand – door crowding.

I really can’t think of a better way to put it. Only on Chicago’s trains have I seen the scenario where so many people on the train crowd near the doors. It’s actually not difficult to see why, though. The seats are laid out pretty badly, with four rows of four seats arranged between the doors that only permit one person to stand in the aisle. Standing here is pretty difficult too, since there are no poles that go from the seat to the ceiling to hold on to (just waist-high bars on the seats themselves). The double seats parallel to each set of doors (see photo above of the woman sitting under the Amstel beer ad for a better idea) don’t make the situation any better, especially because they’re protected from the door by panels which offer people the opportunity to stand next to the doors.

Often the scenario during periods of high demand is that 4-6 people actually make it into the aisle while about 20 crowd around the doors – where people exit and enter. This is most likely because there is too much incentive to stand near the doors and to not move more toward the interior of the car, mainly because of bad seat layout. The new 5000-series rail cars, which are currently only running on the Pink and Green lines, are a slight improvement because there is wheelchair space near one set of doors in each car and hand straps along the aisle to hold on to. Some cars on the Brown line are an improvement over the normal design as well, which have eliminated about 9 seats per car for standing room and have more poles to hold on to.

What I’m surprised about is that the CTA has not gotten rid of the walls perpendicular to each set of doors, which is where people often stand, crowding the doors and making exiting/entering more difficult and slow. I’m sure I don’t need to go into detail about how slow boarding and alighting can clog traffic, especially during periods of high demand when trains need to run on schedule to maintain optimal traffic flow and headway. The CTA could get rid of these “walls” and install a pole directly between the set of doors (in the image below you can see this pole in the center of the floor). The existing seats parallel to each set of doors would be fold-down seats, only to be used as seats during off-peak service:

Strapontin ("fold down seat")

A “strapontin” (folding seat) on Paris Métro Line 7. Credit: flickriver.com

This woman is sitting in a fold-down seat on the Paris Métro. A sign near the seats informs passengers in several languages that the seats are not to be used in periods of high demand. There’s been a campaign by the RATP (the agency in charge of the Métro, some trams, and some regional trains in Paris) to be mindful of the strapontins and when to use them:

Strapontin ad Paris

“He who relaxes during rush hour risks 2 or 3 groans” (it rhymes in French), part of the “Restons Civils” (“Let’s be Civil”) campaign on the Paris transportation system. Credit: Régie Autonome des Transports Parisiens. Click for more ads (in French).

This rule was widely enforced by social norm when I was in France (the ad campaign started while I was living there): You would be reprimanded by a stranger for sitting in one of these seats when the train became crowded (unless it was an elderly or handicapped person).

The new 5000-series rail cars on the CTA do have wheelchair-accessible seating near a set of doors, but the seats are heavy and are not spring-loaded into an upright position, whereas folding seats for normal use would be folded up by default (like in a movie theatre), and must be weighed down by a person. On the 5000-series cars (and CTA buses), these seats are heavy and in the “down” position by default. By making these seats spring-loaded and informing passengers that they are to be used only during off-peak times, there could be more room and space for passengers to exit/enter the train. In tandem with a center-mounted pole and a removal of the walls near the doors, there could be more room to move about the train:

Siemens Inspiro interior

The interior of a new Siemens Inspiro metro car, to be introduced in Warsaw. Credit: Motor Trend.

This is the new Siemens Insprio metro car, which features a center-mounted pole that branches above to offer more grabbing room. There are still protective “walls” next to the doors, but the doors are wider than a CTA train’s, allowing easier movement. There are no fold-down seats. You can also walk between cars, which is a feature of many rail systems around the world (I wrote about it once) and would surely free up more room on trains.

I haven’t had the opportunity to experience the new 5000-series rail cars during rush hour. Hopefully I will have this opportunity soon when they move to the Red line, which I take daily. From what I see, however, people still crowd around the doors and sparsely populate the (now wider) aisles. The doors aren’t wider, meaning people can’t move in and out as freely, and the walls next to the doors are still there. It seems that some small steps have been made, but not enough for an entire new series of rail cars. Some are hung up on the new aisle-facing seats, which arguably free up more space in the aisle. But the crowding near the doors is still pervasive.

If the CTA can’t increase capacity in terms of service offered (perhaps due to budget issues but also because the loop is probably at peak capacity during rush hours, moving 5 train lines), it can at least try to make more of what it already has. And with the new trains costing nearly $1bn, there should have been more thought and testing into what works. But if the members of the Board don’t even ride daily, how can we expect a train design that actually alleviates congestion from the inside?

Rapid Failures: Why Can’t We Walk Between Cars?

This post is part of a multi-part series about the failures of rapid transit in America with a general focus on Chicago, as it is a city I know well and recently moved to full-time.

Something that’s always struck me as bizarre on most American train networks is the inability (or illegality, more accurately) to move between train cars on a subway system. You’ll see this whether you’re in New York, San Francisco, or Chicago: You enter a car and unless you leave to walk to another car, you’re essentially locked into that car – especially if the train is packed.

Something that’s struck me about train cars in other cities around the world, such as some lines in Paris, the Berlin U-Bahn, and tramways in all of those cities is that the train cars are articulated – that is, joined together where there would normally be a space. Articulation in American transit is more common on buses – I’ve been on several articulated or “stretchy” buses, as I’ve heard them called. But never have I been on an articulated train in America.

Berlin U-Bahn

Interior of a Berlin U-Bahn train, whose length you can walk from end to end.

Perhaps one of the biggest reasons for this is because in many American cities, the trains vary in length according to the level of demand. For instance, a Brown Line train running on a Saturday in Chicago may only have 4 cars while on a weekday rush period it has 8 cars (6 during midday). In most European cities, the length of the train does not change, and articulated trains are easier to work with.

But does this mean that articulated trains can’t work on a system that has 4 cars at certain times and 6 or 8 at others? Why not have articulated 4-car sets that you can chain together? I don’t think that the need of articulation is so important that it warrants all new cars, but as I wrote last week, the CTA in Chicago did spend over $1,000,000,000 on new rail cars and the only real upgrades are energy efficiency, a new seating layout, and really important phosphorescent (glow-in-the-dark) strips on the floor (it would have been difficult to just stick those on the floor!). Why not actually completely rethink the way the trains accomodate people?

One reason that articulated trains make sense  is because they can effectively de-crowd the train and spread passengers out. If you stand at the end of a train platform in relation to the direction of the train (i.e., stand at the south end of a platform when the approaching train is heading north), you will inevitably see some cars absolutely full of passengers while other trains will look empty in comparison. This most likely has something to do with station design: If most of the preceding stations had entries at the extremities of the platform, the cars at the ends of the train will probably be more full than those at the center, for example. But by connecting all of the cars of a train together, extra space between cars can be utilized. This could accomodate anywhere between 4-8 extra people standing. It also offers the opportunity for those in more crowded cars to spread out to less crowded cars.

Siemens Inspiro Interior

Interior rendering of a new Siemens Inspiro train, debuting on the Warsaw metro this year.

Articulated train sets also make passengers feel safer by eliminating the loneliness feeling found at off-peak hours on some subway systems. It is dangerous and often illegal to travel between cars on a subway, and the conductor call button in train cars doesn’t cut it for safety – Being able to walk between cars might only offer an illusion of safety, but that may be a deciding factor for some riders.

There might be a reason why we don’t find more articulated trains on American transit systems, such as the tight corners of the loop on the Chicago L, but we already have articulated buses that sometimes make very tight corners as well (I’m not an engineer; someone enlighten me). With the massive amounts of money being spent on new trains, you’d think someone we’d take a newer approach to trains with de-crowding power that isn’t just a new seat layout.

Automating the CTA: A Tried-and-True De-Crowding Initiative

At times it may seem like all I blog about is what Chicago could do better, but I haven’t been writing for a while and I have a few things on my mind.

Paris Métro Line 14

Line 14

When I went to school in Paris, I boarded Line 1 from my stop at George V station and switched to Line 14 at Châtelet. My school was at the Bibliothèque François Mitterand station; the trip took no more than 40 minutes total, including walking and transfer times. What was remarkable about Paris’ Métro system and my commute was that Line 14 is a fully automated and as of this year, Line 1, the busiest line in terms of passenger volume, will also be fully automated.

I really enjoyed the service on Line 14, which was frequent during rush periods – trains often came every 90-120 seconds. The technology behind such a short headway is automatic train control which provides a buffer around the trains for stopping time; the same technology enables the RER A (Réseau Express Régional, or Regional Express Network of trains serving Paris and its suburbs) to also run trains as close as 90 seconds apart.

Now, Line 1 on the Paris Métro has been fully automated without any notable interruption to service. This is remarkable as it is the busiest line on the network, carrying about 725,000 people per day (according to a press release by RATP [link, in French] regarding the line’s automation). The automation process began in 2005 with a contract between RATP (the agency in charge of most Métro operations) and Siemens Mobility and the entire planning, purchasing, construction, and testing process will be complete in a few months. For some perspective, the CTA in Chicago began alternatives analysis studies on the “Circle Line” in 2005 – and we have nothing.

Vive la différence?

I want to go over a few things regarding the automation of subway lines… Since I know Paris as best as I know Chicago (and finding public information about projects in other countries is easiest when you know the native language), I will primarily focus on the automation of Paris’ subway lines when making comparisons.

The first thing I think about when I think about automating subway lines is labor. The drivers must go somewhere once their job becomes obsolete, right? I’m pretty good with math but I’m awful at patience, so instead of waiting for a freedom of information request with the CTA I figure I’d do some of my own calculations: Based off of this CTA timetable for the Red Line (the network’s busiest line), I estimate 175 train trips per weekday – if anyone has a better number, please let me know. And if each voyage from terminus to terminus takes roughly 1 hour and 10 minutes (give some time for signal clearance waits), I can estimate that one conductor drives the entire route 6 times on an 8-hour shift, meaning there are 90 distinct conductors for a 24-hour cycle of the Red Line (If I am wrong, again, please give me better numbers in the comments).

I’m generally on the side of labor unions, but I think that there are certain cases where they are warranted and useful and within those cases, there are certain functions unions perform that are beneficial and others that are detrimental. Concerning transportation industry unions, I feel that the detriments outweigh the benefits (although a lot of the blame can be shifted on low government expenditure on transportation, especially compared with roads – but that’s for another day), especially in the case of automation. Article after article talks about how labor is threatened with automated subway systems. I thought certainly that there must have been a lot of friction between transit drivers for the Paris Métro and the RATP, especially considering that France is heavily unionized and very prone to strike, but it seems like that wasn’t the case, since the drivers for Line 1 were shifted onto other lines, and drivers for Line 14 never existed. Of the 250 drivers who have to leave Line 1 as a result of automation, 40 were promoted to other positions and other drivers were allocated throughout the other 14 lines. [source, in French]

I really wanted to be able to calculate the cost of labor and what the figures are, broken down into cost per voyage, per day, per year, etc… But I don’t have the time to do that research, nor the patience to wait for a freedom of information request to come through. What I do know is that CTA train drivers make $29.65/hour.

Interior of Alstom MP 05 trains

Interior of new Alstom MP 05 trains. Notice that you can walk from one end of the train to the other where each car is separated, resulting in increased standing room. The center pole (which has 3 poles to hold on to) and lack of dividers between the doors and seats means even more room for passengers.

Let me cut to the chase – the automation of Paris’ Métro Line 1 is estimated at 629,000,000 € ($817,071,000 as of September 21, 2012), including new track infrastructure, 49 new Alstom MP 05 trains (at 9.7 million €, or $12.6 million, per train), platform screen doors at 25 stations, and a control center for monitoring the operation of the line.

Bombardier 5000-series rail cars

CTA’s new 5000-series rail cars. Pretty much the same old thing with new seats and lights.

Image

Screens inside Line 1′s new cars, displaying destination, next stops, and travel time to transfer stations.

The CTA recently purchased 706 5000-series rail cars (anywhere between 2 and 8 cars makes up one train in Chicago – but for comparison, let’s say 6 cars = 1 train, as it does in Paris) for $1.14 billion, a cost of about $9.6 million per 6-car train. So the fully-automatic trains in France are a bit more expensive, but they’re fully-equipped for the future. If we ever automated a line in Chicago, we’d have to retrofit the trains for automation.

Paris just automated its busiest Métro line for the less money than the cost of some slightly-better rail cars in Chicago. Those two aren’t entirely equal, but it is interesting to note.

Platform screen doors on Line 13 in Paris

Platform screen doors on Line 13 in Paris

Moving on, many people are concerned with safety and the lack of a driver in automated trains. Paris’ Line 14 isn’t without its incidents, but none of the incidents that occurred during operation of Line 14 were life-threatening or required evacuation. Riding that line every day, I never experienced a delay of more than a few minutes, which is a fairly common occurrence during my daily commute on the Red Line. People are also concerned about emergency evacuations on driverless trains; but think about it this way: A worker in a command center has a much better vantage point of what is going on in the tunnel and inside the train than the train conductor, who has no monitors in his/her cabin and may have trouble navigating a packed train during an emergency. And finally, concerns emerge about people falling on to the tracks, which happens without automated trains anyway. The solution is platform screen doors, which were being installed on Line 1 when I lived in Paris, and were installed on Line 14 from the start. These doors do not open until the train is docked in the station, and the train cannot leave the station until both the platform and train doors are fully closed. The doors can be opened from the track side via an emergency handle, if a person were to somehow wind up on the tracks or an emergency evacuation were necessary.

Come to think of it, the only articles these concerns were expressed in were from pro-labor organizations and couldn’t cite any existing automated subway systems that had a major problem. Make your own inferences.

There would certainly be growing pains with any newly automated system, but I believe that any sensible transit organization would take note from another organization that has successfully built both an automated subway line from the ground up and upgraded an existing, very busy subway line with automatic trains that can run 85 seconds apart – an organization such as the RATP (in cooperation with Siemens Mobility, Alstom, and STIF).

My comparison isn’t perfect. Line 1 in Paris is only slightly above half the length of the Red Line, and France (as many European nations) has much greater public support for public transportation infrastructure projects. But overwhelming praise for Line 14 and anticipated praise for Line 1 as a model for retrofitting subways with automated trains surely says something about the success of these systems. The CTA recently announced a Decrowding Initiative which will add just a few trains to the busiest bus and train lines at the busiest times, at the expense of certain bus routes. This initiative is just a band-aid for a larger problem: Increasing urban populations and more expensive energy combined with lower car ownership rates among youth will inevitably lead to higher future demand for public transportation. How many initiatives will have to be announced until the CTA realizes it has antiquated technology that isn’t ready for the future? Making incremental investments for the most important transportation lines in Chicago means a future where public transportation is reliable, frequent, and pleasant. The CTA should take note from other cities that are successfully preparing for their future.

I welcome your comments and corrections below.

Update: I didn’t mention New York City’s anticipated automation of the L line. The line isn’t done and has been marred with difficulties. It wasn’t a suitable comparison for a simple analysis of automated transit networks.