Metro Vancouver Transit Discussion II

[aberdeen5698]

Originally Posted by logan5

It’s pretty easy to predict these vehicles will use the roads far more efficiently when there are enough of them on the roads. Road space won’t be a problem.

It's not that you can change the paradigm when "enough of them are on the roads", you can't really do that until you've eliminated ALL of the manually driven vehicles. And I suspect that's going to be big political obstacle for a long way into the future.

And the idea that a line of autonomous vehicles will somehow be able to turn themselves into something resembling a train is also a fallacy. There needs to be a margin of safety to handle failure scenarios. Blowouts, obstacles falling onto the road, wayward pedestrians or animals, all require vehicles to maintain space around them to give them time to react to unexpected situations. The only way to get train-like spacing is to have a grade separated right of way into which no potential obstacles are permitted to intrude. We have that already - it's called Rapid Transit.

And even if you could run cars bumper-to-bumper, they are STILL far inferior to a transit vehicle in terms of road space per person. Closer vehicle spacing may be able to move more people in cars, but it will never be able to move all the people and it will never be able to do so at the cost and scale of transit.

[logan5]

Originally Posted by aberdeen5698

It's not that you can change the paradigm when "enough of them are on the roads", you can't really do that until you've eliminated ALL of the manually driven vehicles. And I suspect that's going to be big political obstacle for a long way into the future.

And the idea that a line of autonomous vehicles will somehow be able to turn themselves into something resembling a train is also a fallacy. There needs to be a margin of safety to handle failure scenarios. Blowouts, obstacles falling onto the road, wayward pedestrians or animals, all require vehicles to maintain space around them to give them time to react to unexpected situations. The only way to get train-like spacing is to have a grade separated right of way into which no potential obstacles are permitted to intrude. We have that already - it's called Rapid Transit.

And even if you could run cars bumper-to-bumper, they are STILL far inferior to a transit vehicle in terms of road space per person. Closer vehicle spacing may be able to move more people in cars, but it will never be able to move all the people and it will never be able to do so at the cost and scale of transit.

Ok, how close together, in your estimation, could autonomous vehicles get, while going 60km/h (and maintaining an acceptable level of safety. 5 feet? 10 feet? 15 feet?

[cganuelas1995]

Originally Posted by WarrenC12

Flexibility. Automated buses could use transit only tunnels in dense urban areas then spread out on their own routes on roads outside of it.

And I reckon emergency vehicles might be able to get a use out of them as well

Originally Posted by CanSpice

Why can't buses being driven by humans do that?

Well, for starters, a bus driven by a human needs a human that is trained to operate said bus, and that human needs to be compensated for it, and they need to be compensated fairly. And because they are a human, they are vulnerable to things like injury, illness, and death, as well as plain old occasionally being unable to be present and punctual to operate said vehicle, usually as a result of the things they are vulnerable to.

Originally Posted by WarrenC12

Imagine a ton of autonomous buses driving around, various sizes. There are no fixed routes and times. You open up the app say "I'm here and want to go there". The system works it's magic and says a bus will be there in x mins to take you and drop you at your destination (or maybe a "station" in a busy downtown) and will cost xyz. It's less convenient than Uber and will take longer as it picks up and drops off others, but that will be reflected in the cost.

You mean like a very small vehicle, one capable of transporting a passenger and the passengers belongings?

[WarrenC12]

Originally Posted by cganuelas1995

You mean like a very small vehicle, one capable of transporting a passenger and the passengers belongings?

I was thinking of a mix of vehicles, everything from larger buses down to single occupant vehicles.

[Migrant_Coconut]

Originally Posted by cganuelas1995

And I reckon emergency vehicles might be able to get a use out of them as well

We already have bus tunnels. The question then is whether or not the cost/benefit makes them more worth it than bus lanes.

[cganuelas1995]

Originally Posted by WarrenC12

I was thinking of a mix of vehicles, everything from larger buses down to single occupant vehicles.

So like pods, sedans, vans, minibuses, full and articulated buses? Any cargo options? I feel this can be really fleshed out

[Changing City]

Originally Posted by cganuelas1995

So like pods, sedans, vans, minibuses, full and articulated buses? Any cargo options? I feel this can be really fleshed out

There is a Roads Fantasy thread, and a Transit Fantasy thread. Autonomous vehicles in tunnels sound like they belong in one of those, rather than the Transit Discussion.

[aberdeen5698]

Originally Posted by logan5

Ok, how close together, in your estimation, could autonomous vehicles get, while going 60km/h (and maintaining an acceptable level of safety. 5 feet? 10 feet? 15 feet?

This paper seems to suggest intervals of 0.6s to 1s or distances of 7 to 25 meters. At 60 km/h, you travel almost 17m/s, so the intervals work out to something like 10 metres or more.

It's worth noting that during times of congestion in bumper-to-bumper traffic the following distances of manually driven cars are already well below that.

[ilikeredheads]

I feel like some of you do not drive on a regular basis or have never been behind the wheel if you really think fully autonomous driving on public roads is going to be a thing in the near future.

Real world traffic is a highly uncontrolled environment, with lots of variables that can happen unexpectedly. To have a system that can react to all these variables without being a liability is way more difficult than you think.

[Migrant_Coconut]

The bigger problem is that the idea of personal rapid transit itself is flawed - it's all the downsides of transit and driving combined with none of the benefits.

Tracked guideways like Morgantown's are a bust. Podcars like Heathrow's don't seem to be much better. More recently, the Boring Company has tried a variant on the podcars in Vegas, except underground and more expensive... and has more or less just reinvented underground gridlock.

If the premise itself is no good, then automation will just be polishing a turd.

[logan5]

Originally Posted by aberdeen5698

This paper seems to suggest intervals of 0.6s to 1s or distances of 7 to 25 meters. At 60 km/h, you travel almost 17m/s, so the intervals work out to something like 10 metres or more.

It's worth noting that during times of congestion in bumper-to-bumper traffic the following distances of manually driven cars are already well below that.

From that same paper.

...capacity could be doubled or tripled at urban intersections (Kockelman et al., 2016; Lioris et al., 2017), and
quintupled on freeways...

[s211]

Originally Posted by ilikeredheads

I feel like some of you do not drive on a regular basis or have never been behind the wheel if you really think fully autonomous driving on public roads is going to be a thing in the near future.

Real world traffic is a highly uncontrolled environment, with lots of variables that can happen unexpectedly. To have a system that can react to all these variables without being a liability is way more difficult than you think.

I watched a Phoenix Waymo conduct its way at a busy threeway stop and pedestrians not giving a fcuk and waddling everywhere, and the car was flawlessly safe. It was pretty cool to watch it adapt on the fly.

[CanSpice]

Meanwhile, in Metro Vancouver, the only self-driving transit we're going to have in the next twenty years is SkyTrain, so maybe all of this autonomous transit and autonomous vehicle talk should go in some other thread?

[logan5]

Originally Posted by CanSpice

Meanwhile, in Metro Vancouver, the only self-driving transit we're going to have in the next twenty years is SkyTrain, so maybe all of this autonomous transit and autonomous vehicle talk should go in some other thread?

Link to source that says that.

[Changing City]

Originally Posted by logan5

Link to source that says that.

TransLink are just buying the replacement trolley fleet intended to operate for the next 20 years that won't be designed with any autonomous operating systems. So a significant part of the bus network certainly won't be autonomous.

The battery buses they're buying so far also don't have potential for autonomous operation. If there's successful trial somewhere (other than in China), maybe it will be part of TransLink's future plans, and vehicles capable of autonomous operating will be bought. (They cost quite a bit more, obviously, so they're unlikely to buy them until there's a reality of using them). That's not currently the case.

[WarrenC12]

Of course Vancouver will be far behind the curve. Look at how long it took us to get Uber.

Lots of people seem to have a hard on for traditional buses, that's fine. I take them as an option of very last resort.

[Changing City]

Originally Posted by WarrenC12

Of course Vancouver will be far behind the curve. Look at how long it took us to get Uber.

Lots of people seem to have a hard on for traditional buses, that's fine. I take them as an option of very last resort.

That probably explains why you don't seem to understand how difficult automating our bus service would be. It's unlikely we'll have extensive dedicated transit only lanes. If we did, and the technology ever becomes viable, then robocabs would be very limited in where they could pick up and drop off passengers.

In the meantime I can't imagine how a robo#20 bus would operate with a stroller already on board, a Chinese senior with limited English wanting to board, and a mobility scooter with an overweight disabled male trying to get off and needing the ramp, but 2 Danish students with big backpacks blocking the aisle. In reality the driver sorts that out and there might be a bit of slippage to the schedule.

You would probably have to replace the entire road transit and private vehicle infrastructure with a range of smaller, specialized and autonomous cabs. That's going to be very expensive, a lot of people wouldn't want it, and in the meantime it belongs in a fantasy thread.

[WarrenC12]

Originally Posted by Changing City

That probably explains why you don't seem to understand how difficult automating our bus service would be.

Nowhere did I say it would be easy. You're constantly reading things that people didn't write to fit your own narrative.

[CanSpice]

Originally Posted by WarrenC12

Nowhere did I say it would be easy. You're constantly reading things that people didn't write to fit your own narrative.

That's great, it would be hard financially, it would be hard politically, it would be hard logistically, it would be hard technologically, and it would be hard socially. What isn't hard is moving all of that discussion to the fantasy thread.

[logan5]

100% automated transportation is going to be hugely disruptive to public transit, and relatively soon. This is not fantasy at all. We have a working system in place in San Gran, LA and Phoenix.