5 ms·
On very busy or very long escalators, standing is probably faster for everyone. However, most escalators aren't that long, aren't that busy and are a fair bit s
by Tharkun 8y ago
On very busy or very long escalators, standing is probably faster for everyone. However, most escalators aren't that long, aren't that busy and are a fair bit slower than walking. In which case everyone should ideally be walking. But because lots of people have luggage or aren't comfortable walking up escalators, the whole walk left/stand right thing feels like a pretty decent compromise.
What the world really needs is more stairs for those who are able, and more lifts for those who have difficulties with stairs. Or maybe multi-lane, variable speed walking escalators, but I guess those are sci-fi...
- darkpuma 8y ago> "On very busy or very long escalators, standing is probably faster for everyone. " How so? I don't see how an escalator could possibly be slower than everybody standing (discounting the possibility of somebody actually walking backwards.) > "Or maybe multi-lane, variable speed walking escalators, but I guess those are sci-fi... " I've badly wanted to try that ever since I read Caves of Steel. It sounds terrifying but exhilarating.
- dx87 8y agoYou can probably safely fit more people on the escalator if everyone stands still, so while individuals would be going slower, overall throughput on the escalator would be higher.
- darkpuma 8y agoSomebody walking on the elevator does not inhibit the person behind him from standing still. The person who stands behind a walker is going to get on and off the elevator at the same times.
- zamadatix 8y agoThis thought experiment about 2 people makes sense on a short, unfrequented, escalator. The conversation is on long busy escalators however and someone will almost always be sandwiched between two people in that scenario. If not there wouldn't be any concerns about throughput considering one lane is already used for walkers.
- darkpuma 8y agoThe article and studies it's discussing are about the scenario where standers avoid the left side of the escalator because they think it's reserved for walkers. But what I'm saying is that allowing walkers to walk whenever there is opportunity while also telling standers to use both sides gives you identical throughput to everybody standing (while allowing any walker who gets the opportunity to get off the escalator faster.)
- mac01021 8y agoIf walkers are allowed to walk on the left whenever noone is standing in front of them, then I will naturally tend to avoid standing on the left. Unless the left is already blocked by other standers.
- darkpuma 8y agoI'm not sure how to break this to you, but walkers can also walk on the right when nobody is standing in front of them. I've been doing it my entire life and not once has anyone stopped me.
- eikenberry 8y agoThe classic latency/throughput trade-off.
- darkpuma 8y agoIf everybody steps onto the elevator at a set rate, momentarily stands still, then some of them begin walking forward iff there is space to do so, how does that cause reduced throughput? To be clear, I'm not talking about some sort of "walk on the left" convention, which plainly reduces throughput. I'm talking about the "walk if you want to no matter what side you're on" convention, which seems to be far more universal.
- rvp-x 8y agoif a queue forms to get on the escalator
- darkpuma 8y agoWalking on elevators does not cause more cue than standing on them. Two people can step onto an escalator and if the one in front then takes a step forward because the one behind was inside his personal space, the one behind is not impacted. Furthermore since stepping onto an escalator necessarily involves walking up to that escalator, if the minimum distance between people walking were causing a throughput limitation that would manifest whether or not people on the escalator were walking.
- mcny 8y agoIf they really want to "fix" this "problem", they can try to make the escalator narrower so there is only one "side" possible and fit more escalators. You shouldn't carry bulky luggage in an escalators anyways.
- darkpuma 8y agoOr they could just turn up the escalator speed a bit. A mere 10% speed increase would probably make a big difference, but I suppose there are almost certainly regulations on how fast an escalator can run.
- Tharkun 8y agoI've wondered about this as well. Different places seem to have different escalator speeds. It's been a while since I last saw someone who was too afraid to get on an escalator, but I do frequently unsteady elderly folks who struggle a bit. I guess that's the demographic that's preventing faster escalator speeds.
- xbmcuser 8y agoWalking does cause a larger queue as people get on one side (non walking) of esclator mostly . So 1 person gets on instead of 2. I have now started standing in what is considered the walking lane once people see that I am not going to climb they got on behind me and the throughout of the elevator increases.
- tinus_hn 8y agoIf everyone is standing still they can be closer to each other so more people can be on the escalator at the same time, so more people can make the trip in the same time, just like how driving at a slower speed prevents congestion because the drivers have to keep less distance. This has of course been measured. If the escalator is packed the people on it can’t start walking because there is no space. And escalators are often packed in stations, for instance when a train arrives and a thousand passengers get off the train at the same time.
- autokad 8y agoyeah, I'm also tired of lets force everyone to do something to speed up the mean, when what we really care about as anything in this universe is the tails. some people dont need to rush to make their next flight, but some do.