4 ms·
I don't think use of the Underground will increase by 60% by 2050. Mass transit will be a bad memory by then. When our driverless cars head underground, they w
by stretchwithme 11y ago
I don't think use of the Underground will increase by 60% by 2050. Mass transit will be a bad memory by then.
When our driverless cars head underground, they will all be going at the same high speed.
- ptaipale 11y agoWhatever the speed, they won't have the capacity that mass transit provides in things like London underground.
- Silhouette 11y agoThat depends very much on the design. Remember that at any given time, most of the tunnel in the Underground is empty. Most of the time you're waiting in stations is because other people are getting on or off, not you or your own party. And most of the stations you stop at (and travel slower before and after) aren't relevant to your own journey. A network of high speed paths with efficient out-of-path access for starting and ending journeys could in principle offer dramatically higher throughput than current mass transit on the Underground, even with individuals or small groups sharing private vehicles. The trouble is that the amount of time and money required to build that network underground would be staggering even under ideal conditions, and conditions under London are very far from ideal.
- Symbiote 11y agoBut the bits of tunnel that aren't empty are extremely full. A tube train carries over 1000 people maximum. But it also takes up about 1km of track before the next train (estimated since stations are about 10 minutes walk apart). We have 1 person per metre of tunnel. The platform is roughly the length of the train, 130m. A small cart carrying one person could be, say, 1.5m long, and take up 3m (very optimistic, I think). Now we only have ⅓ person per metre of tunnel. Let's say 200 people get on or off the train at a busy station. There's a train every 120 seconds. The additional space is slightly larger than the train (the platform). If it takes 10s to get off the cart, in 120s we can load/unload 12 carts. But they're smaller, so we can do, say, 75 in the existing platform space (130m). That's 900, so this bit isn't a problem. Maybe it adds up; I'm actually far to tired to continue working it out...
- Silhouette 11y agoSorry, but I don't think some of your assumptions make much sense. Tube train tunnels are large. If we were developing some alternative method of transport, why would we reserve the entire cross section of current tunnels for single occupancy vehicles? As demonstrated by the Tube itself, much higher densities could be achieved, even with private vehicles, either by having multiple people per vehicle or by using smaller vehicles that didn't require as much space and could run alongside others. Of course density isn't the only thing that matters. Consider that the average speed of an Underground train is about 20mph. A fit person can run faster than a Tube train moves over short distances.[1] Clearly if you had a system where loading and unloading happened without blocking the main tunnels, significantly higher speeds could be achieved over the kinds of distances a typical passenger might be travelling. [1] No, really! https://www.youtube.com/watch?v=PH_Z8Ghuq6E https://www.youtube.com/watch?v=PH_Z8Ghuq6E
- lorenzhs 11y agoTube tunnels relay are not large. They vary a bit by line but in general they're really really small and severely limit the design of new trains. There's no way you'd fit two vehicles side by side in those tunnels, and they're not high enough to stack them either. Nor are they interconnected well enough to make one the passing lane. Have a look for yourself if you get to London, the entire tube system is mind-boggling.
- Silhouette 11y agoI'm from the UK and have been riding the Tube for decades. If we're talking about a hypothetical alternative transport system to be built underground and carry private vehicles potentially with single occupancy, you could probably fit four such vehicles within a cross sectional area not much bigger than the existing tunnels. The engineering practicalities would presumably still favour a round tunnel, which would be less efficient in space if you were running smaller vehicles in parallel, though. Then again, as I mentioned elsewhere, the cost of actually doing any of this in both time and money would be wildly prohibitive today and seems likely to remain so for many years to come, so this is all rather hypothetical anyway.
- apsec112 11y agoDigging a single tunnel under the length of London (the Crossrail project) cost ~$25 billion. And that's just one tunnel. London has a population of ~8.5 million, and a single road lane with automated vehicles can handle around ~3,000 cars per hour. Multiply it out, and the cost of tunnelling a road system large enough to carry the entire city at high speed becomes mind-bogglingly huge.
- stretchwithme 11y agoI said "WHEN our driverless cars head underground". I did not say all the travel would be underground. Small single passenger vehicles can also pass each a lot more easily in the amount of space a train takes up. You could have two levels in the amount of vertical space a train takes up. I also don't think a train takes up less of the available space and time than individual cars would, when has to stop at all the stops. All other trains must wait too. The flow is constantly interrupted. Robotic underground construction is also going to be a lot cheaper. Whatever happens and whatever the mix between walking and vehicle transportation evolves, people are going to prefer to travel in their own vehicles over being crammed into cars with strangers and having to navigate escalators. This particular system is shaped by the current technological constraints, not so much by people's desires. When technology advances, it will eventually change.
- fredoralive 11y agoI don't think you've seen a London Underground deep tube line, you ain't getting two levels of cars in that space. No overtaking either, as they use separate tubes for each direction. Even the larger, shallower subsurface lines are UK loading gauge, which isn't big enough for double decker trains. I doubt any double decked traffic on existing UK railway land without extensive expensive rebuilding. As for "robotic underground construction", it's not like they have a load of Navvies digging by hand nowadays, a modern tunnel boring machine is pretty damn robotic, but it still costs a fair chunk of change to tunnel.