Is it inefficient to walk up the escalator?

A study in London found 74.9 per cent of people choose to stand instead of walking, especially on the longer ones. With this ‘stand on the right, walk on the left’ rule, we’re giving up 50 per cent of the space on our escalators for roughly 25 per cent of our commuters.

Look for this problem next time during rush hour where the “standing” side of the escalators ends up with a line of people trying to get on. It may seem counterintuitive, but people who are walking up escalators to save seconds off their commute are actually slowing everyone else down.

Efficiency aside, there’s another reason why walking on escalators might be a bad idea—safety. Escalator accidents are much more common than you think.

A CBC investigation found that escalator accidents happen every second day in the Montreal Metro. In the U.S., about 10,000 escalator-related injuries end in emergency room visits every year.

Many of those victims were likely walking. A  study in Tokyo found almost 60 per cent of escalator accidents between 2013 and 2014 resulted from people using escalators improperly, which includes people walking or running on them.

Here is the full story, via Michelle Dawson.  Walking up the escalator remains time efficient, however, if those choosing to walk have much higher valuations of time than those who choose to stand.  Might that be the case?

Comments

and forcing people who want to walk to stand would also be pareto inefficient, of course

Steve Landsburg addressed this escalator issue at length last year on his blog.
See here: http://www.thebigquestions.com/2018/12/29/escalators-the-geeky-version/
here: http://www.thebigquestions.com/2018/12/25/escalating-matters/
and here: http://www.thebigquestions.com/2018/12/19/ups-and-downs/

Respond

Add Comment

Exactly. A certain percentage of the population can't handle walking up an escalator without injuring themselves, so they stand.

Everybody wins.

Respond

Add Comment

Respond

Add Comment

Sure, running on escalators is foolish like a Trump voter but walking is improperly using them?

Good One!

Or, like a liberal standing on a broken escalator waiting for someone to make it move.

Respond

Add Comment

An escalator can never break: it can only become stairs. You should never see an Escalator Temporarily Out Of Order sign, just Escalator Temporarily Stairs. Sorry for the convenience.

Millenials will never know his genius

As a millennial, I totally listened to him in college. Drove to CT to see him. I'm guessing you don't know what age bracket a millennial is.

Respond

Add Comment

Respond

Add Comment

I was at an airport once and the "up" escalator stopped. It was a gradual stop so people were still perfectly upright and for a second or two kept standing still, wondering what to do. They then did what most of us would do, they started walking up what now seemed like a staircase.

But as the people walked up the escalator, their weight caused the escalator to move downward. Which was not the direction they wanted to go so they started walking faster.

This sounds like the set-up for a humorous scene in a Buster Keaton movie, but not everyone was able to move their feet real quickly and real well and a woman tripped and fell. She wasn't hurt, but this had clearly turned into a hazardous situation.

By that time some airport person had shown up and told the people on the escalator to just walk down the escalator and take the stairs up.

So that broken escalator was not equivalent to a stairway, it was more dangerous. I presume that escalators have the equivalent of a parking brake, and maybe they're even supposed to kick in automatically (elevators became hugely safer when someone [Otis?] invented emergency brakes for them), but on this escalator the parking brake did not kick in automatically.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Context matters.

At the mall: walking is more efficient
5 Avenue/53 St station: walking is definitely more efficient

Meant to say...
At the mall: standing is more efficient
5 Avenue/53 St station: walking is definitely more efficient

You mean because the escalator never works?

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

I have argued that the expectation that the groceries be bagged slows down everybody in the absence of a bagger to assist the cashier, as his scanning will have a lower duty cycle. In that mode, self bagging like in discount grosery store is better.

I would rate the claim that some people walking up the escalator slows down everybody as probably true. However, if the average right side stay still dude is like the average airline passenger who slows down everybody as s/he puts hand luggage in the overhead bin, the claim might become shaky.

> I have argued that the expectation that the groceries be bagged slows down everybody in the absence of a bagger to assist the cashier, as his scanning will have a lower duty cycle.

In my experience, people are much slower at bagging their own stuff than the cashier (me included). So it might end up with the cashier waiting for you to finish.

It might depend on other factors. If the shop is small and crowded enough, it may be more efficient to get rid of the shoppers faster by bagging for them.

At my local Kroger, they do the two lane checkout. One lane for the previous checkout to bag their groceries and another for the current checkout. It works well in practice. The people who write checks are the worst and yes they still exist in the world's largest economy.

>yes they still exist in the world's largest economy.

In fairness, the laws governing checks are more consumer friendly than credit cards, which in turn, are more consumer friendly than other electronic transactions.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

"some people walking up the escalator slows down everybody"
Surely it doesn't slow down the walkers. This logic only applies if there are more standers than capacity allows. Many times there is no queue to get on the escalator.

The queue forms when the plane or train arrives, and both are scheduled to all arrive inefficiently in short spans of time through the day with many hours of little activity.

During these times, many of those queued have cargo that is most easily handled on flat floor at a width that is almost the width of the elevator. Some of these people do the natural thing of walking straight onto the elevator and stopping because their luggage can't roll. But that blocks those trying to walk up the elevator.

Others rearrange themselves to reduce their width, but that requires going slower and taking up more space on the elevator, forcing those behind them to slow, causing congestion. That's what creates the queue at the elevator.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

I'm told that this is why there are no signs telling tourists in DC to walk on the left, stand on the right: it would open Metro up to liability for escalator accidents.

Respond

Add Comment

If Americans are as obese as the statistics say, then there's no way you can walk past anybody on the escalator.

Respond

Add Comment

If you need to get somewhere fast, don't take the escalator. Just take the stairs next to it. You can go way faster.

How do you figure? Surely on an escalator you'd go as fast as on the stairs plus the speed of the escalator?

Unless, of course, the escalator is going the other way.

Speed is a vector. If the escalator is going the other way, its speed has a negative sign.

Admittedly been a long time since diff eq and calc 3.

I thought velocity was the vector.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

People block the way. Some people don't move out of the way. Even if they move, they're slow to do so. I take stairs when in a hurry.

Clearly you've never been to London. Brits are very serious about their queuing. (Or their standing on the right.)

You'll be asked to stand to the side very soon.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

WTF? Coase, Coase, Coase: Why don' those immobile buggers walk, nay run, up the escalator? :-)

When will luggage come with legs and feet instead of wheels, so the luggage can run up the escalator with you?

The only time I encounter escalators is either where their are travelers, transportation hubs, or shoppers, multistory malls. A high share of people have bulky bags, some bulky enough to have wheels.

Respond

Add Comment

Respond

Add Comment

Sure, the total number of people going up the escalator/minute might be lower. However, allowing people to self-select into the "prefer to get there faster" and "prefer to stand and not exert the energy" might mean higher aggregate utility for escalator rides. AKA - people who stand get slightly less utility by having to wait in a line, but people in a serious hurry get to travel the distance more quickly.

You mean, pick between taking the full width of two steps or take half the width of 3-4 steps.

Never seen anyone hoist their luggage, or kid(s), up over their head to get on the elevator taking less space than walking in a wider walkway.

Respond

Add Comment

Respond

Add Comment

This is making assumptions about "commuting" and "leisure." Say I go to a restaurant I take a table but order right away, eat very quickly and pay immediately. How does that negatively affect ANY other diners? And how could that possibly have a negative effect on the people waiting for a table?

Respond

Add Comment

Shouldn’t we outlaw standing on the escalator if we’re trying to maximize efficiency?

Yes. This article gets the math wrong.

In a crowded subway station (eg any subway station at rush hour) the escalator forms a severe bottleneck, with people queuing up at the bottom.

Walking on an escalator dramatically increases throughout, by 100-200 percent, thereby dramatically alleviating the bottleneck and resulting in increased efficiency for everyone.

It’s the standers who are causing dangerously overcrowded platforms and slowing everyone’s commute down.

"Walking on an escalator dramatically increases throughout, by 100-200 percent"

Standing 50cm apart on the escalator and travelling at 5km/hour
Versus walking 150cm apart and travelling at 10km/hour

There would be more throughput on the standing side.

Sure, if you assume walkers take three times the room, it is less efficient for anyone to walk. But I don't think it is true. First, most people are on their own and don't like standing next to a stranger, so the escalator will normally have only one person per step. If they all stand to the same side, the unused half can be utilized by walkers.

That's a theory, but the experiments at Holborn Tube Station that started all this found that the speed/spacing ratio was such that throughput was faster without the no-standing-on-the-left rule.

I have seen repeatedly at a DC Metro station that with both rows walking it's fastest. Things slower but still OK with one row walking and the other standing. But some standers want to stand in the walking row! Then things break down.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

"There would be more throughput on the standing side."
Perhaps, if one falsely presumes a constant flow of people to the escalator. Yet this simplifying assumption is not true. If we take the situation where a train arrives, typically the first people to arrive at the escalator will be same people more likely to walk up.

Second, looking at overall throughput is likely a simplification. By allowing walking, we allow the people that care about a getting out of the station more quickly, and are able to walk up stairs, to move through more quickly.

I seldom see a queue on only the standing side. Perhaps the heuristic would be to split until a sufficient queue forms on the standing side that completely filling the escalator with two standers on each step would be possible.

Are you accounting for the width of the escalator being typically half the width of the walkway approaching it?

Apply your theory to a two lane road going from 24 feet wide to 16 feet wide.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Well, thanks for pointing that out. The math indeed looks funny in the post.

Nevertheless, DC has a good point: the experiment did point to an increase in the capacity with the stand-only rule.
https://www.theguardian.com/uk-news/2016/jan/16/the-tube-at-a-standstill-why-tfl-stopped-people-walking-up-the-escalators

DC seems to conclude that people choose to be farther apart when walking. That does not match my personal experience in Brazil, the US or Europe.

I believe more plausible explanations may be that (i) walkers are blocked from getting to the escalator by standers crammed near the access point or (ii) standers are still waiting in queue after all walkers are long gone and the left-side escalator real estate is idle.

Respond

Add Comment

Respond

Add Comment

Yes, we should flog people who are not moving with a sense of purpose.

Respond

Add Comment

Respond

Add Comment

file this in the "really stupid questions" folder...zero value

Economist much?

The situation is a simple and natural scenario for analysis with broader implications. It is similar, if not identical, to having HOV lanes on highways.

Respond

Add Comment

Respond

Add Comment

This is a great example illustrating the difference between the economist’s notion of efficiency (maximize value) and the engineer’s notion of efficiency (maximize quantity / size / weight etc.), a frequent cause for people talking past each other.

Yup. Similarly, when mathematicians solve games of fair division, their notion of "fair" often differs from economists'. Economists are more likely to try to take into account potential heterogeneity in people's preferences and to try to allow trading and side payments, depending on what the game set-up is.

Respond

Add Comment

Respond

Add Comment

At busy times, at least at my metro station, it self organizes with people standing on both side, at less busy times the left side is left clear.

Agreed. The vast majority of the time, most escalators aren't full to the point where a line with any kind of serious delay exists at the bottom. If it did, then surely one of the first steps would be for the people in line to "stand" who want to go faster to convert to "walk" instead...

Respond

Add Comment

Respond

Add Comment

Who cares? This is why Trump will get reelected. You libs can suck it again in 2020! #maga2020

Respond

Add Comment

At a slightly higher level of Zoom having a rule simply gives users of a system the choice of self-selecting into different options with "perceived higher or lower efficiency" and widens the range of use cases for the system as a whole. In other words, I can use the system in 2 modes: for a leisurely trip where I am optimizing for distraction (say reading MR on my phone while standing still on the escalator) vs. optimizing for time (say when I'm late for a meeting and my mind is racing to come up with plausible excuses!) At the margin, this option can affect decisions of whether to rely on the tube for all trips. What's optimal for the micro system may not be optimal for the macro system.

Respond

Add Comment

Having walking side on half of escalators would be fair?

Respond

Add Comment

"Walking up the escalator remains time efficient, however, if those choosing to walk have much higher valuations of time than those who choose to stand. Might that be the case?"

Sounds entirely plausible to me. If the people standing are in a big hurry, they could just switch to walking. Whether I'm a stander or a walker on a given day depends mostly on what my schedule that day is.

Respond

Add Comment

Nice spinmeistering. Obviously, if the goal is to avoid "slowing everyone else down", then everyone should walk up the escalator. It doesn't take a genius to recognize that the capacity of the escalator is maximized when everyone walks. Having said that, it's perfectly reasonable to make special accommodation for standers, even though it sacrifices capacity.

I would call this piece preemptive spinmeistering. I'm sure the author recognizes the obvious truth that walking is faster than standing. To preempt any complaints about standers blocking walkers, the author spins a story where, ackshually, it's the walkers that are slowing down the standers.

Let's see, standers take 4 steps on the elevator while walkers take only two?

Or do walkers need 4 steps in the elevator while standers only need two?

By spending half the time on the elevator taking up a window of 4 steps you increase the throughput of standers taking 2 steps, because the elevator throughput is higher at 1/2t × 4 steps than at 1t ×2 steps.

Respond

Add Comment

Respond

Add Comment

"those choosing to walk have much higher valuations of time than those who choose to stand"

Some people have an inflated view of their real worth. That's why I pay my workers no more than minimum wage even if they think they deserve more. If you think you can make more then leave I tell them. So few take me up on my offer. Suckers.

This handle and the other with the blatant anti-Semitic stuff.

anonymous/mouse is officially over the edge.

Respond

Add Comment

Respond

Add Comment

Additional utility to walking: exercise, especially when you're going to your desk job; signaling- I'm fitter than the other slugs on the train and have somewhere important to go. Standing: a few more moments of not being where you don't want to go.

Respond

Add Comment

I suspect the escalator is a metaphor for the economy: not everyone travels at the same speed in the economy, those who move slower holding back those who move faster. Cowen might suggest that the fastest movers have resolved the problem by building their on fast lane.

Respond

Add Comment

They actually experimented with restricting walking in some stations at peak hours, but it failed as an experiment, mostly due to the utility argument that Tyler cites. The people who want to walk up really value the additional speed, whereas if you're happy to queue to stand going up them clearly you do not value the additional time savings. This manifested when they tried the experiment in having a few really irritated commuters, who were also much more likely to complain.

Respond

Add Comment

This is why TfL tried to ban walking on the escalators at my local tube station. But the experiment failed, because even with staff standing at the bottom of the escalator to remind people, few people were willing to stand on the left and generally everyone was annoyed.

Respond

Add Comment

Just one more aspect of US societal insanity, along with beige wall-to-wall carpet, lawns, circumcision, football, pet worship and catch-and-release fishing. The time differential between walking up the escalator and standing on it is less than insignificant in any context. Is the time that can supposedly be saved placed in a time piggy bank to be shaken out later to extend a vacation in Toledo? The obsession with meaningless amounts of time is a mental illness.

Patience, once considered a virtue, is now non-existent. A large share of the blame for this sad fact goes to US retailing, where the interval between the initiation of desire for a product and its delivery has been made more important than its quality or price, for a competitive advantage. Witness the success of Amazon.

Additionally, individual hurry demonstrates, at least to the individual, a self-importance that belies the fact that in any rational view their lives are of no importance to anyone else. "Get out of my way, I'm important and I need to be somewhere." This attitude is particularly evident on the highways, a situation that would astonish and repel normal people only a few generations ago. Worse yet, this attitude has itself become normal, at least in the Americano culture, forcing those that cling to a reasonable pace in life to adopt franticism or be considered as standing in the way. It's a sad state of affairs that doesn't receive the attention and analysis that it deserves.

So much ridiculousness in your post.

It's easier to believe people are mentally ill than to believe that their actions reveal their preferences?

With respect to transportation, saving a couple of seconds could save you a 20 minute wait for the next bus or train. Saving seconds could mean the difference between being late for work or not.

I do agree though that lots of drivers on the roads don't really make up much time with their weaving and speeding, putting everyone at risk for almost no gain. Those risks are substantially lower with pedestrians.

Mentally ill people often reveal that fact by their preferences.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Something of potential interest to a tiny minority of people who work in whatever engineering-related trade encompasses design of movable and immovable staircases.

Respond

Add Comment

If I'm standing it would probably be due to my bad knee. Forcing me to walk because you think your time is more valuable than my pain is not good

Respond

Add Comment

Would anyone stand shoulder to shoulder with another unknown rider? I believe if there was no walking it would be one at a time due to social norms on proximity/space between unknown people. People walking by doesn’t seem to violate this norm.

You wouldn't stand on successive steps either. So if walking is banned, people just alternate.

Respond

Add Comment

Respond

Add Comment

This was looked into closely in London a couple of years ago, maybe the most detailed article is this one:

https://tech.labs.oliverwyman.com/blog/2018/08/08/escalators/

It's a close-run thing, depends on the height of the escalator, and many of the articles seem to confuse concepts. The number of people on the escalator at any one time (bandwidth) is irrelevant. The system should optimise for throughput (number of people through station).

Respond

Add Comment

Walkers need more room, standers can pack in. The net cost is apportioned by the packages being carried. Moderate loads, will walk tr stand, since marginally speed and effort are balanced. This marginal 'price' will keep the queues stable. (They simulate Gaussian independent arrivals). When queue variance equals queue mean. At that point, there is no observable advantage to either line, in aggregate.

Another abstract algebra tree, the escalator is the tree trunk, the world of random walks is you hologram.

Street traffic works the same way, a sort of breakthrough in understanding.

When the wiat queues meet the stability condition (mean equals variance), then the street traffic appears like Gaussian arrivals. That means, cars between intersections are maximally independent. That is, there is room in the lanes for the minor shuffle needed to move the flow. The hologram effect, drivers know what speed prevents crashes at the next stop. They can calibrate by just knowing how long the queue is in typical intersections.

Consider the drivers manual, what is it?
An instruction sheet on how to manage traffic congestion, almost everywhere in the manual. The new driver is being hologramized to the typical vehicle sizes, a small set. He is being taught an axis of symmetry, car separation along the linear road. That is the hologram in reverse, the DMV rules are making the constrained intersections always properly quantized. Still, the abstract tree applies.

Too bad it's totally insane

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

An otherwise unasked question, apparently:

how many of these commuting geniuses on their escalator are interacting with their mobile devices on the mechanical stairway?

Mobile device consultation by motor vehicle drivers in transit DOES slow down the traffic flow for ALL motor vehicle drivers in each locality where mobile device use by motor vehicle drivers is permitted, whether by law or by the absence of enforcement options.

Pedestrians consulting their mobile devices impede pedestrian traffic flow at least as much, if mobile device use by pedestrians is anything nearly as ubiquitous today as motor vehicle drivers consulting their mobile devices while they attempt to drive.

Anti-social technology at work: no complaints at all.

(What kinds of consumption efficiency are driven by motor vehicle drivers who insist on impeding motor vehicle traffic flow by using, misusing, and abusing their "mobile device privileges" while driving? How much extra carbon spews into our atmosphere each day now that mobile device use encourages mindless driving strategies?)

Poor driving performance resulting from the advent of mobile device use by motor vehicle drivers has been a feature of contemporary life for an entire quarter century already. Driver response times are down at intersections, in lanes where these rude idiots routinely drive five- or ten-miles-an-hour BELOW posted speed limits (regardless of left-lane/right-lane customs and regulations), in driving much too close to vehicles ahead of them, et cetera, et cetera, et cetera.

What a startling coincidence that the threats posed by the advent of Technogenic Climate Change should emerge almost simultaneous to the advent of anti-social technologies that only encourage thoughtless carbon spewing (has ANY economist dared to measure just how far driver response times have been reduced cumulatively because rude idiot drivers are incapable of operating motor vehicles efficiently while consulting their mindless "mobile devices"?)

Mobile device manufacturers and service providers owe us all an explanation that they are not offering to provide.

Economists owe us some overdue explanations, too.

"What a startling coincidence that the threats posed by the advent of Technogenic Climate Change should emerge...."

Speaking of which, what did you think of Nature retracting the major ocean warming paper a few days ago? Amature climate scientist Nicholas Lewis had found errors in the Princeton groups paper a year ago, who admitted the problems last November, so it is interesting that Nature took almost a year to retract it.

I have gained no impression over the past year that climatic phenomena are paying strict attention to the data reported of them.

Nevertheless, I plan to continue to monitor monthly satellite displays of sea ice extent in what yet remains of our two sub-sub-tropical polar regions.

Oceans and seas have absorbed enough carbon, it seems, to have begun decreasing oxygen levels commonly being measured to depths of one hundred meters. Food chains not being disrupted yet by the marine dissemination of microplastics are being impaired already by oxygen deprivation in the planet's oceans.

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Respond

Add Comment

Even if no one walked up the elevator, many (most?) people wouldn't crowd side-by-side on an escalator.

Respond

Add Comment

"Many of those victims were likely walking" I'd guess most injuries are from getting on and off the escalator. Walking would reduce the transition and possibly decrease injuries.

Respond

Add Comment

Ah, no. Those who stand reveal a lower value of their time than those walking. Alternatively, the walkers have some motive other than time to walk that the standers dont.

Each of the standers has the choice of walking, hence they aren't "slowed" at all.

Moreover, a stander is only slowed when the escalator reaches maximum congestion which is rare.

While some people have enhanced utility from standing or walking on escalators, many people CHOOSE to use stairs, revealing their preference for exercise or personal space.

One should question the inefficiency of having non-moving stairs between up and down escalators rather than relieving congestion with more escalators.

Respond

Add Comment

It is more of a fast-lane/slow lane system.
If there's no one standing in front of you, you walk on the right, so people that have to run for a train/meeting/whatever can pass you on the left.

How many escalators cover so much rise the time on the escalators exceeds 45 seconds?

"They found that walking up the escalator took 26 seconds compared with standing, which took 40 seconds. "

Actual data from engineers studying escalators.

Why You Shouldn’t Walk on Escalators https://nyti.ms/2ozotQV

Finally googling the topic.

Let's see, walking on escalators reduces throughput while increasing average travel time, saving walkers only 13 seconds.

Plus, the uneven load on the mechanism causes escalators to break down more frequently. I guess forcing everyone to walk the escalator means walking on escalators is superior to standing?

Respond

Add Comment

Respond

Add Comment

Lazy people cause themselves to wait. Why not get rid of the escalator? We’d all be better off burning a few more calories and save money too.

Respond

Add Comment

It's perfectly possible to get injured by an escalator while standing still. As a kid, an escalator grabbed the end of my shoelace and ground along until it got a piece of my toe. If I'd been walking, I probably would have noticed earlier and escaped injury...

Respond

Add Comment

The analysis is wrong because it fails to take into account the other bottlenecks of subway travel, namely the turnstiles or access gates. The initial stock might be the people waiting to access the escalator and the respective flow might be the rate at which those people go from the first step to last. However this creates a second stock of people at the end of escalator which must then flow through turnstiles, into trains, or onto a walkway. The last is likely to have the most turbulence as people walk at different paces once off a smoothly flowing machine, but the previous two could bring about trouble, maybe not so much the train, but there is a tendency of bunching at the escalator and surge when the train arrives. This could cause additional injuries. The real issue would bet the rate at which people could get through the turnstiles, the second flow of this system. If the first flow at the escalators becomes more efficient "increases" without a proportional increase in the flow of the second juncture, back ups will ensue. The more the disparities in the flows, the faster the back ups will accumulate. The escalators don't have sensors to turn them off once of platforms on which the stocks inhabit reach maximum capacity, causing a higher potential for injury.

Respond

Add Comment

Huh? People walking up the escalator only slow down people unwilling to walk up the escalator. Anyone who finds the added wait time unpalatable may move their ass a little and get on up there. On the other hand, people wanting to stand in line for the opportunity to stand on the escalator are in fact just getting more standing time in, which seems to be what they were after in the first place, so it's a perado good.

Respond

Add Comment

Walking up escalators is an element of my general fitness practice in addition to saving time. It also includes taking the stairs rather than elevators, walk commuting, and, playing sport.

Respond

Add Comment

I use the metro regularly in an EU capital city, and I have my doubts about the capacity increase envisioned by those who want to abolish the walking lane. The fact is, people do not want to fill the escalator to capacity. It's rare for two strangers to stand side by side, even at the busiest times. It is also quite common for there to be an empty step between riders, even on the right side. People just don't want that much closeness. Besides themselves, people are often carrying things, sometimes big things. I would suspect the maximum practical capacity of such escalators is closer to 50%, with alternate steps occupied on alternate sides.

Respond

Add Comment

I think he best way to analyze this is to consider throughput and average wait time for each type of escalator user and overall. Allocating 50% of capacity for 25% of users could still improve those numbers as long as it increases escalator throughput enough and thus decreases average queue wait time overall.

For example, assume both the standing and walking sides of the escalator are fully utilized and standers spend 20 seconds on the escalator, while walkers spend 5 seconds. In this case, the 50% of the escalator used for walking will have 4 times the throughput of the walking side. Right away, that means more people can potentially traverse the critical resource in less time.

The next consideration is what happens with average queue length for each type of elevator user. First, consider the case where there's a queue for both the standing side and the walking side. For this case, the wait time on the walking side will be considerably less than the wait time on the standing side. As long as walking increases throughput enough to offset the increase in wait time caused on the standing side by allocating half of capacity to walkers, the average wait time overall will be reduced. OTOH, even with a potentially lower average wait time overall, the worst case wait time for the standing side will be higher. So standers pay much higher costs (in wait time) even though more people can traverse the resource in less time overall (a potential fairness issue).

This is the type of problem that freshman Computer Science students are often asked to model using a simple computer program. They'd use a random number generator to create "people" wanting to use either side of the escalator and track queue lengths and average wait time under various assumptions for arrival rates and escalator traversal times.

And thanks to google, here's the result of this type of analysis:
https://www.telegraph.co.uk/news/2016/03/22/the-simulation-that-proves-standing-only-escalators-work-on-the/

Respond

Add Comment

I agree. It’s still risky to get up on the platform and get off it, especially when the ladder moving under you disappears - clothes may get stuck in the gap. An escalator, for example, tore off a child’s thumb in Singapore and strangled a drunken sushi cook when the edge of his sweatshirt fell into the gap between the stairs and the landing platform. Of course, there is an alternative: the good old staircase. But almost 12 thousand people in the United States die every year after falling from the stairs. It seems that moving up or down is always risky. Good luck!

Respond

Add Comment

That is, ASSUMING that the escalator is currently saturated. If it's not crowded, then the escalator throughput will increase by providing a standing lane since in order to make space people will have to occasionally take one step back, but walkers can take multiple steps forward and there are fewer pairs than walkers (since the escalator is so uncrowded, people who get on at the same step become rare).

If it is crowded, then in practice there will be some pair of people blocking the escalator and forcing it into the everybody-stands-high-throughput mode, so "stand on the right, walk on the left" doesn't have an effect anyway.

IOW, it's good to keep the "stand on the right, walk on the left" norm, as long as you recognize that it'll be broken on crowded escalators and that's a good thing.

Yes, I think this is the best analysis. Most of the answers (not just in these comments, but seemingly in the experiments and simulations too) do not account for the heterogeneity of escalator use and users.

Using your examples, plus some more: is the escalator saturated?

Is there a queue, and is it for just the walking side or for everyone?

How often does the escalator become congested (at an airport this might happen every time a plane lands; at subway stops it's probably every time a train arrives but maybe only during rush hour depending on how busy the station is; at a department store big floods of passengers are going to be less common)?

How many passengers will inevitably be walkers (probably a high percentage at airports where people are encumbered with luggage and a smaller percentage at subway stations during rush hour when most passengers are commuters who at most have a briefcase or knapsack)?

How many passengers are in a hurry (probably a lot during commuter rush hour; probably a smaller percentage at airports -- but the ones who are in a hurry and in a really desperate rush to catch their plane so we have to also pay attention not just to the percentage of walkers on the escalator but also how high their time cost is)?

As others have mentioned, on most escalators that I've seen it's rare for two standers to share the same step so I am skeptical about the alleged higher capacity of standing-only escalators. This is especially true at airports where people have luggage that they'll almost always want to put down, either beside them on the step, or on the step that's ahead of or behind them.

There might be an analogy here with zipper merges, which traffic engineers have discovered are superior to having everyone make an early merge into the soon-to-be-single lane. I do believe in that research result. But I'm skeptical about the experiments that claim to show that everyone-stands is a superior policy, unless and until they've accounted for the heterogeneity of escalator usage and users.

Respond

Add Comment

Respond

Add Comment

And then you have riding/walking DOWN the escalator, which is going to require its own extensive set of studies

Respond

Add Comment

It seems to be the efficiency is that 50% of the escalator is being used by 25.1% of the people not that walking up the elevator is inefficient. So why is the conclusion that the people walking up are slowing the process instead of the people who are waiting to stand aren't slowing things? To me it sounds like if 50% were standing and 50% were walking it would be faster as the space usage efficiently allocated similar to 100% standing plus you have the added benefit that half of the elevator is spending less time in that space. Also to get to the 50/50 split you need to change 24.9% of the peoples behavior which seems more fair than changing 25.1% of peoples behavior.

Respond

Add Comment

Respond

Add Comment