Calgary Public Transit II

[dazzlingdave88]

I thought this was interesting. Definitely has some exaggeration to it, but he does bring up some good points.

http://www.humantransit.org/2015/11/self-driving-cars-a-coming-congestion-disaster.html

[sim]

Originally Posted by milomilo

But why should this automatically be a bad thing? There is no need to pursue a goal of mode share percentage just for the sake of it. If we are transporting more people than before with the same infrastructure, who cares?

Well I think the answer, in part, to your first question is in your second one. Our existing infrastructure could, in fact, be used a lot more efficiently with a mode shift. In addition, and along the same lines, from an environmental perspective, shifting away from vehicle use is desirable no matter how well utilized vehicular transportation infrastructure is.

But those seemingly inconvenient but salient issues aside, your point is valid and it would be a small, but important step in the right direction. At the moment, increasing connectivity in the street network is everyone's favourite transportation planning principle. In fact, you basically just have to say "increases connectivity" and recommendation justified. (Unfortunately, it's a poorly understood concept and metric on good days).

Anyway, the point is that connectivity is intended to do just that - use more or less the same amount of lane kms, but organize them in a way that distributes traffic more evenly among roads, ergo, better utilization and more peak spreading. In principle, this is a good objective for multiple reasons. Unfortunately, this is probably the most difficult thing to implement in existing communities. You'd have an easier time politically "encouraging" mode shift (by not continually building interchanges and widening roads for one), than punching new street links through communities. But perhaps more importantly, it is directly contrasting with how our suburbs have been purposely built for some time now - namely to rely on a heavy hierarchical network to direct traffic away from residential areas and onto a few key roads.

Aside from the irony that doing so actually creates more of the very thing those types of street network structures are trying to avoid, on the surface of it, it's a very rational thing for anyone living anywhere to desire - less noisy, dangerous traffic going through their neighbourhood. So until this very poignant issue can be reconciled with what you are saying, which is in principle valid, there is a very legitimate reason for wanting mode share changes - because it can't [be reconciled]: Lots of traffic sucks, and it sucks from everyone's viewpoint until they are that traffic.

[CalgaryAlex]

Originally Posted by dazzlingdave88

I thought this was interesting. Definitely has some exaggeration to it, but he does bring up some good points.

http://www.humantransit.org/2015/11/self-driving-cars-a-coming-congestion-disaster.html

That is bang on. All self-driving cars will do is make driving safer. I can't see that technology revolutionizing transportation in a city. Rapid transit methods carrying hundreds of people in the same vehicle to their destination will still be far more economical and environmentally-sound than anything else.

[DizzyEdge]

Originally Posted by CalgaryAlex

That is bang on. All self-driving cars will do is make driving safer. I can't see that technology revolutionizing transportation in a city. Rapid transit methods carrying hundreds of people in the same vehicle to their destination will still be far more economical and environmentally-sound than anything else.

I could see self driving cars tapping into traffic heat maps and reducing some congestion by taking clear routes.

[You Need A Thneed]

Honestly, self driving cars could greatly increase the capacity of existing roads. Cars that could communicate with each other don't need nearly as much space in between them, and theoretically, we wouldn't need traffic lights, or interchanges.

[ByeByeBaby]

Originally Posted by You Need A Thneed

Honestly, self driving cars could greatly increase the capacity of existing roads. Cars that could communicate with each other don't need nearly as much space in between them, and theoretically, we wouldn't need traffic lights, or interchanges.

I'm fascinated - when and where are you posting from? I'm from Calgary in 2015, a place where pedestrians occasionally cross roads, where tires blow out and loads fall off pickup truck beds, where children and dogs chase balls near roads, where my damn cell phone sometimes can't load a web page, and where self driving cars represent 0% of the operating fleet.

[You Need A Thneed]

Originally Posted by ByeByeBaby

I'm fascinated - when and where are you posting from? I'm from Calgary in 2015, a place where pedestrians occasionally cross roads, where tires blow out and loads fall off pickup truck beds, where children and dogs chase balls near roads, where my damn cell phone sometimes can't load a web page, and where self driving cars represent 0% of the operating fleet.

All the technology needed to deal with that is pretty much available now. Implementation is the tricky part.

[milomilo]

Distances between cars will increase if anything - computers will drive to the book and allow for the minimum safe stopping distance, something human drivers don't do. See also automated rail systems. Even in a completely controlled environment the systems always allow for the train behind to leave enough distance to come to a complete halt even if the train ahead stopped dead.

[MalcolmTucker]

But you reduce reaction time and some systems include coordination, so safe stopping distance can take into account simultaneous breaking of many 'platoon' cars.

[milomilo]

Originally Posted by MalcolmTucker

But you reduce reaction time and some systems include coordination, so safe stopping distance can take into account simultaneous breaking of many 'platoon' cars.

All just fantasy right now, AFAIK no automated trains utilise a platoon system even though their environment and maintenance is far more controlled.

There will be a very long transition period between the time self driving cars are allowed in mixed traffic and when there are enough on the road to justify making sections self driving only. And only then could you even think about these things suggested, which still probably won't be allowed.

[dazzlingdave88]

Originally Posted by milomilo

All just fantasy right now, AFAIK no automated trains utilise a platoon system even though their environment and maintenance is far more controlled.

There will be a very long transition period between the time self driving cars are allowed in mixed traffic and when there are enough on the road to justify making sections self driving only. And only then could you even think about these things suggested, which still probably won't be allowed.

I agree, this isn't the magical cure to congestion, especially if it increases the number of people driving. If 20% more people drive because of it, it would have to improve driving efficiency by 20% to even break even.

The transition period will be long, because of affordability.

Also, are these cars going home after they drop people off, or are they going to be parked downtown?

[fusili]

Originally Posted by You Need A Thneed

All the technology needed to deal with that is pretty much available now. Implementation is the tricky part.

I can't wait to jaywalk with impunity when we get to full adoption of driverless cars.

[para transit fellow]

One thought experiment predicts even greater congestion. No one has explained where all these automated cars are going after they drop us at our destinations

The challenge was something like:

You take your automated car to work and send it home.... (hmmm.. automated car is ameliorating parking congestion but all the inbound cars now create another rush hour)

When your automated car returns to take you home, you decide to run an errand. With no parking available, you instruct the car to continue circling the block for half an hour while you attend to your errand. (hmmm...Automated car has added traffic congestion)

The example is not perfect but I am not sure we will automatically see a decrease in traffic congestion.

[sim]

Originally Posted by You Need A Thneed

All the technology needed to deal with that is pretty much available now. Implementation is the tricky part.

Exactly. This is probably the ne plus ultra of how smartphone waving technophiles always believe that technology can simply solve everything. Technology does not exist in isolation. If this technology really were at a level (and maybe in the near future it will be) that it is reliably safe enough so that people can walk out in front of vehicles without worry, a much large social issue will be forced to play out: Essentially streets will either finally become true public domain, or we'll finally be forced to truly concede them to the almighty vehicle. That is, if we actually want to see the supposed congestion reduction benefits / improved capacity utilization they promise (intersections aren't going anywhere soon):

Originally Posted by milomilo

Distances between cars will increase if anything - computers will drive to the book and allow for the minimum safe stopping distance, something human drivers don't do. See also automated rail systems. Even in a completely controlled environment the systems always allow for the train behind to leave enough distance to come to a complete halt even if the train ahead stopped dead.

And the promise is real. It's been pretty well proven that they will be able to maintain closer spacing. This is really about the easiest of all problems to solve in this realm. See Volvo's conveys in Spain for a working example. There is a lot of psychology involved in car following - this would all be removed.

Train signalling systems are also moving to communications. The difference with trains is that they are much, much more efficient and therefore, much more difficult to stop quickly at comparable speeds/environments (metal on metal is your friend). Trains are rarely in comparable environments - they travel quicker or are often in tunnels. It's not an overly meaningful comparison.

But all that being said...

Originally Posted by para transit fellow

One thought experiment predicts even greater congestion. No one has explained where all these automated cars are going after they drop us at our destinations

The challenge was something like:

You take your automated car to work and send it home.... (hmmm.. automated car is ameliorating parking congestion but all the inbound cars now create another rush hour)

When your automated car returns to take you home, you decide to run an errand. With no parking available, you instruct the car to continue circling the block for half an hour while you attend to your errand. (hmmm...Automated car has added traffic congestion)

The example is not perfect but I am not sure we will automatically see a decrease in traffic congestion.

... this isn't just a thought experiment. More than one study has shown that vehicle kilometres travelled (vkt) will increase, not the least, the report posted which prompted this discussion. The more interesting question, in terms of congestion is whether the improved capacity utilization will/can offset the additional vkt. I'm not sure this has been studied (or studied well), but in any case, even if it has, the additional vkt aspect is a very human factor, thus, not overly susceptible to being predicted (well) or linear.

Which really brings it full circle, even if congestion were to be reduced, we still haven't addressed the underlying problem - a reliance on an intrinsically inefficient form of accessibility. In fact, we'll probably have made it worse.

[mwalker_mw]

Self-driving cars will not be the revolution - only a stepping stone. It is shared self-driving cars will be the revolution.

A car that picks me up wherever I am and drops me off right at the door of where I need to be and the goes on to pick someone else up without ever parking is extraordinarily efficient compared to what we have now. I would speculate even if average speeds were to halve for safety and balancing load (outside of freeways) there would still be a net gain in people getting where they want to be faster on average.

The technology will be here in 10 years. It is only a question of the will to implement because humans are terrible at accepting communal and averaged benefits.

That being said, I think in our increasingly connected world physical presence is becoming less and less important and if we are freed up to 'connect in' to our destination with the majority of our attention while the car does the navigating the level of acceptance for communal efficiency will start to win out from the current 'I NEED TO BE THERE RIGHT NOW EVERY OTHER DRIVER ON THE ROAD MUST DIE DIE DIE DIE AND GET OUT OF MY WAY AND DIE' attitude. Ten minutes late for a meeting doesn't matter so much if you can connect by mobile phone, see everyone via videoconference and seamlessly wander in a few minutes later having missed nothing and delayed no one. Fifteen minutes behind on picking up little Johnny is not a big concern if he/his minders can see your location and predicted arrival time on their phones down to the second.

There are huge segments of the worlds population that travel exclusively by trains in their day-to-day activities. Shared self-driving cars, from a user perspective, are nothing more than trains that happen to locate themselves to you rather than the other way around. Maybe we don't call them self-driving cars - perhaps 'microtrains' instead. Being beholden to the schedule of a shared resource becomes rather palatable if the delays are not significant. No one complains about subways that run on 3 or 4 minute headways and I suspect microtrains could provide a similar level of service in dense areas once broadly deployed.

If we look at the potential savings from (nearly) eliminating car crashes and the associated property damage and medical costs (not to mention the secondary administrative costs of insurance companies and lawyers) there is suddenly a lot of money that could go towards navigation infrastructure for intelligent transportation networks.

[Socguy]

I've been following the automated cars discussion with some interest. I really don't have a clue what they're going to mean for the future of transportation. I can see the arguments for both sides. Still, I don't think that we should alter any plans for expanding transit right now as expecting some new technology to save us would be folly.

Although, if the predictions in this video are correct, transportation might be solved simply as a side effect of automation. https://www.youtube.com/watch?v=7Pq-S557XQU People simply won't have as many places they need to go to.

[VIce]

Originally Posted by mwalker_mw

Self-driving cars will not be the revolution - only a stepping stone. It is shared self-driving cars will be the revolution.

A car that picks me up wherever I am and drops me off right at the door of where I need to be and the goes on to pick someone else up without ever parking is extraordinarily efficient compared to what we have now. I would speculate even if average speeds were to halve for safety and balancing load (outside of freeways) there would still be a net gain in people getting where they want to be faster on average.

The technology will be here in 10 years. It is only a question of the will to implement because humans are terrible at accepting communal and averaged benefits.

That being said, I think in our increasingly connected world physical presence is becoming less and less important and if we are freed up to 'connect in' to our destination with the majority of our attention while the car does the navigating the level of acceptance for communal efficiency will start to win out from the current 'I NEED TO BE THERE RIGHT NOW EVERY OTHER DRIVER ON THE ROAD MUST DIE DIE DIE DIE AND GET OUT OF MY WAY AND DIE' attitude. Ten minutes late for a meeting doesn't matter so much if you can connect by mobile phone, see everyone via videoconference and seamlessly wander in a few minutes later having missed nothing and delayed no one. Fifteen minutes behind on picking up little Johnny is not a big concern if he/his minders can see your location and predicted arrival time on their phones down to the second.

There are huge segments of the worlds population that travel exclusively by trains in their day-to-day activities. Shared self-driving cars, from a user perspective, are nothing more than trains that happen to locate themselves to you rather than the other way around. Maybe we don't call them self-driving cars - perhaps 'microtrains' instead. Being beholden to the schedule of a shared resource becomes rather palatable if the delays are not significant. No one complains about subways that run on 3 or 4 minute headways and I suspect microtrains could provide a similar level of service in dense areas once broadly deployed.

If we look at the potential savings from (nearly) eliminating car crashes and the associated property damage and medical costs (not to mention the secondary administrative costs of insurance companies and lawyers) there is suddenly a lot of money that could go towards navigation infrastructure for intelligent transportation networks.

People signed up for Car2Go even though it only misses your concept in terms of automation. The idea would be to sell the use of a vehicle as a service, as opposed to property someone wants to own.

[mwalker_mw]

Originally Posted by VIce

People signed up for Car2Go even though it only misses your concept in terms of automation. The idea would be to sell the use of a vehicle as a service, as opposed to property someone wants to own.

Certainly Car2Go is a great proof of concept for the shared ownership model (or rather, the subscription model) that is an essential part of this shared, automated future. I think all the pieces are starting to make themselves known for a revolution in transportation but I'm uncertain what the tipping point will be or when it will happen.

For me it can't be soon enough - I need all the productive time I can get... (or, at least some time to nap...)

[McMurph]

As a user of car2go I have always assumed that a car share model will be the end game when it comes to urban self-drive cars. I imagine Car2Go with an Uber type interface... good-bye taxis, car-share and a lot of private car ownership.

The prognostications of privately owned self-drive cars trolling around the city are weird and, I hope, a little alarmist. I know there are a lot of people who have no clue how successful car share works (I've got relatives from the eastern US who have taken photos of those "great little car-things" here) -- I wonder if the doomsayers about self-drive cars are just lacking in imagination. If not, then I hope to hell that a technical / legislative barrier to trolling private vehicles is developed before such a problem becomes widespread.

[dazzlingdave88]

I wonder if automakers are worried about this future?

I know with Car2Go and living near work, my wife and I have went from two vehicles to one. That one vehicle barely gets 20 000 km/year. And that is with all of our family living either in BC or in Northern Alberta.

I would love it if self-driving cars were introduced on a subscription model. I love paying for a vehicle only when we need it.

I think a lot of times people romanticize the journey inside of the goal of transportation. Status symbols to get us from point A to point B.

Just a question to get people thinking. I posted this a couple of weeks ago to my Facebook page. I really think this technology will help immensely for safety.

"If Teleportation were invented tomorrow with the caveats:
- 1.24 million people would die from it each year
- Those deaths would cost $518 billion
- It would cost $1 trillion dollars globally to operate/maintain
- It would be the leading cause of deaths in people 15-29

Would we accept this new technology?

These stats are from worldwide motor vehicle use (the best I could find). 14 people were killed on Alberta roads this weekend. As most of you know, this subject impacts me greatly. My mom was one of those statistics a couple of years ago, but I refuse to let her be just that. There needs to be a serious debate between convenience and safety.
Why does society deem it acceptable for someone who endangers or kills people while operating a motor vehicle to retain their license in the majority of cases, but if an airline pilot exhibited those same behaviors they would lose their license?
For every one person who dies from a plane crash, 1629 die in car crashes. All of this and I haven't even mentioned the 50 million people who are injured worldwide each year in car accidents."