Metro Vancouver Transit Discussion

[Spork]

I think that the anticipated capacity for SOV autonomous vehicles is a bit ambitious with the first generation. The mentioned 2 foot distance between vehicles would never fly with a skittish public - SkyTrain's buffer area is much larger than that, and those vehicles are already automated and communicate with each other (albeit indirectly, so perhaps latency is an issue).

[aberdeen5698]

Originally Posted by Spork

I think that the anticipated capacity for SOV autonomous vehicles is a bit ambitious with the first generation. The mentioned 2 foot distance between vehicles would never fly with a skittish public - SkyTrain's buffer area is much larger than that, and those vehicles are already automated and communicate with each other...

This is what I just don't get about autonomous vehicle advocates - they seem to have these very high expectations of space savings due to automation, yet they're ignoring all of the safety and reliability issues that this implies. Automatic train control in protected rights of way is a mature technology, but even the most aggressive systems rely on a "moving block" exclusive occupancy that guarantees enough distance for a following vehicle to come to a safe stop in the event of a failure.

Yet we're somehow expected to believe that autonomous vehicles in a much harsher threat environment are going to magically be able to zip around with merely a few feet of separation and whiz through intersections in an interleaved pattern at full speed? That the propulsion, braking and sensor systems these privately owned vehicles will rely on to operate safely will never fail? That these will all operate at 100% efficiency in inclement and slippery weather?

I can envisage this sort of thing happening some day, but nowhere near as quickly as a lot of people seem to believe.

[WarrenC12]

Originally Posted by logan5

Labour for taxis is 66% of the cost, so it will be muxh cheaper for autonomous taxis.

Source?

[ryanmaccdn]

Originally Posted by WarrenC12

Source?

Downtown to New West is 60$ ..... gas alone in a hybrid probably 8$ of that.

[WarrenC12]

Originally Posted by ryanmaccdn

Downtown to New West is 60$ ..... gas alone in a hybrid probably 8$ of that.

How long is that trip? 45 mins? Implying labour is 66% of $60 = $40 for 45 mins = $53.33/hour?

If you want to state numbers as facts please back them up.

[Genauso]

The taxi drivers operate like sharecroppers. They pay to rent the vehicle per hour, and if they don't bring in enough revenue they will lose money. The owner of the taxi license takes most of the money, with actual operating expenses from maintenance, fuel, and labour being minor share. The license holder in turn has paid the local city for the monopoly right on operating fare based rides. Ask your driver the next time you take a ride for the actual numbers, it's been a while since I have asked.

As for New West to downtown Vancouver, one way it's normally 20km / 30 min (less if they're speeding during a late night trip). A Prius has 4.7 - 5.1 L/100km fuel efficiency.

If that trip is $60, subtract the fuel and it's about $58.50 for a 30 minute trip. There's depreciation, maintenance, insurance, etc. There's a chance of an empty cab return trip, but the principal argument that we have instituted a system to keep prices elevated by blocking new entrants holds.

Personally I have taking a taxi as my least preferred option in Vancouver, and it has nothing to do with perception of cleanliness or anything like that. There are cities where a taxi is my first choice. Once I order a cab in Vancouver, I have no idea when it will arrive and then I have no idea what the fare will be. It follows a formula, but one that is too complicated to seem fair as you just have to accept the number on the box that seems too high. There's also the issue of what payment methods they might decline, or what destinations they aren't interested in serving. On top of that we have so many different jurisdictions, with all kinds of rules and barriers that don't let a willing driver give a ride to a willing customer at any price.

==========

Back to Translink. I don't worry about the busiest routes being able to justify their costs and surviving if ridership declines for any reason. SkyTrain, SeaBus, any bus route that is normally standing room only.

I do worry about two areas on the horizon of the next 20-30 years:
- investing in new routes, hoping that it will create its own demand later
- people who can't drive, this will increasingly be the elderly, and who simply can't afford what we have chosen to price taxi fares at

For the fringe bus routes, we can definitely get more butts in the seats if the bus doesn't have to follow a predetermined route and if it can signal riders when it will arrive to eliminate time wasted waiting (the best ridership is in areas where it's only 5 minutes or less until the next train/bus, gambling on a 30minute+ extra wait on each leg of the trip is a huge deterrent.)

For the elderly and disabled we have the HandyDart system. I know very little about it, but I'm vaguely aware that for eligible riders it operates like a minibus/shared taxi except it must be booked weeks in advance. I'm also vaguely aware there is an issue with decline in quality for that service to do with cutbacks and outsourcing. Regular buses aren't a great place for the elderly today if it's crowded, there are sudden stops, they have to wait outside during all weather, if they have groceries, and if they have to walk far on a hill from their stop for example.

There is a huge opportunity for TransLink to take its weakest point and innovate.

Take a look at Finland, after all they do almost everything better. We have no excuse to not try by pretending something that exists is impossible.
Motherboard: Choose Your Own Route on Finland's Algorithm-Driven Public Bus
Written by BRIAN MERCHANT
October 15, 2013
Wired: New Helsinki Bus Line Lets You Choose Your Own Route
BY KEITH BARRY 10.11.13

Helsinki Times: The future resident of Helsinki will not own a car
04 Jul 2014

When the Mayors' Council put out their proposal for the investments in the next 30-50 years of TransLink I wasn't just disappointed, I was disgusted. They proposed what would have been a bad plan if they were living 30 years ago. There was zero ambition for self-improvement at TransLink, it was just new spending to solve new problems. Otherwise it at best planned to just maintain what has proven to work, and at its worst ignored current failings. There is no consistency or metrics of success, other than their own pride and importance.

[cabotp]

Originally Posted by logan5

Why do you say that? Do you think its not technically possible? If mode share is even between car and transit, then road capacity would only need to double. Every indication says this is easily achievable. People keep saying it can't be done, but no reasons why it can't be done.

Well we are talking about automated vehicles replacing transit. So we can't have a mode share that is even between car and transit.

And I've stated why it can't be done. It is a simply problem of 2D physics. You can't take a bus load of people and put each of them in there own automated vehicle and end up using the same area of the road. It simply is physically impossible. And even if you take 50% off the bus and put each of them in their own automated vehicle. You still end up using more 2D physical space than what you would of with that one bus.

Now automated vehicles are a good invention. Just that they will never replace transit as a whole.

[logan5]

Originally Posted by cabotp

Well we are talking about automated vehicles replacing transit. So we can't have a mode share that is even between car and transit.

And I've stated why it can't be done. It is a simply problem of 2D physics. You can't take a bus load of people and put each of them in there own automated vehicle and end up using the same area of the road. It simply is physically impossible. And even if you take 50% off the bus and put each of them in their own automated vehicle. You still end up using more 2D physical space than what you would of with that one bus.

Now automated vehicles are a good invention. Just that they will never replace transit as a whole.

A study by Columbia University. Although this is theoretical freeway capacity gains, the same rules would apply for city streets. There is a lot of wasted space on our roads.

Self-driving cars, formerly a science fiction fantasy, are growing increasingly closer to reality. A fleet of Google-modified autonomous vehicles have logged over 300,000 miles with no accidents, reports the Atlantic.

The new study was written by Columbia Unviersity's Patcharinee Tientrakool and is set to be presented at a Institute of Electrical and Electronics Engineers conference this week.

Tientrakool found that safe vehicle distance could shrink to about 16 feet for robo-cars going 75 mph, compared to over 115 feet necessary for safe stopping by human-operated cars at the same speed.

http://www.nydailynews.com/autos/aut...icle-1.1153419

[theho]

The thing about theoretical capacity / throughput increases that self driving cars might promise is that they require a system where every actor on the road is a self driving car coordinating with all the rest. On a highway maybe. In an urban area - no way is it reasonable to assume that with things like people crossing the road or bicycles. As well as totally chaotic things like raccoons, cats and loose dogs.

I think that self driving cars are going to be great - but mostly in terms of solving the last mile problem of getting between home or work to the nearest station. Mass transit is still going to be the most efficient way of moving people over longer distances in an urban area. Everyone using a fleet of self driving cars that do not need to / cannot park in the downtown in the morning would just mean that we would have crazy rush hour congestion in both directions.

[theKB]

Originally Posted by theho

In an urban area - no way is it reasonable to assume that with things like people crossing the road or bicycles. As well as totally chaotic things like raccoons, cats and loose dogs.

These systems are already in production cars that prevent collisions with those sort of issues that present themselves in normal city driving.

https://www.youtube.com/watch?v=U7kl1-FKKYk

[theho]

Originally Posted by theKB

These systems are already in production cars that prevent collisions with those sort of issues that present themselves in normal city driving.

https://www.youtube.com/watch?v=U7kl1-FKKYk

Right. What I was contradicting without clearly indicating (sorry) was the idea that driverless cars have efficiencies in terms of road surface usage and speed significantly over human operated vehicles. (I expect them to behave more or less like a really good human driver) For example an idea that with driverless cars is that you do not need traffic lights and therefore can route vehicles through an intersection with very tight timings and reduce unnecessary delay. Those models do not include pedestrians or bicycles also using those intersections.

What I meant by the chaos that raccoons and other things introduce I am not saying it because the cars would hit them, but because with a system with very tight timing - any unplanned delay (a braking event) would possibly have very far reaching effects literally chaos in a mathematical sense. (Imagine a system where everything is running at millisecond near misses. A unplanned delay might mean that large sections of the network effecting thousands of cars would have to slow down to accommodate that event)

[Porfiry]

Originally Posted by logan5

A study by Columbia University. Although this is theoretical freeway capacity gains, the same rules would apply for city streets. There is a lot of wasted space on our roads.

Highways are the best case scenario for automated vehicles. Surface roads in a city are the worst case scenario. It's not honest to pretend the gains will be equivalent.

[logan5]

Originally Posted by Porfiry

Highways are the best case scenario for automated vehicles. Surface roads in a city are the worst case scenario. It's not honest to pretend the gains will be equivalent.

Obviously I disagree. A person driving on a freeway has a lot less variables to deal with, you just drive in a straight line. City streets have traffic lights and cars going in all different directions. There's more potential efficiency improvements for city streets. Even if you have full size cars with a 6 foot buffer between them running in platoons, you would still get 4 times more capacity per lane.

Anyways, at this point I'm just repeating myself, so we'll have to wait and see how Googles fleet of 100 fully autonomous vehicles work out when they begin testing them in an urbanized environment early next year. These vehicles will have no steering wheel, or break pedal, so Google must have some confidence in what their vehicles are capable of.

[Kisai]

Originally Posted by Porfiry

Highways are the best case scenario for automated vehicles. Surface roads in a city are the worst case scenario. It's not honest to pretend the gains will be equivalent.

Highways in FLAT terrain are the best case scenario. This is where there are no obstructions for GPS/cellular networks, and no constant weather/speed/grade shifts. Automated vehicles along the West coast of North-America, especially in cities is not likely never going to happen in any significant capacity unless every car can communicate with each other.

The "Google cars" are designed for city use only. They literately have a 3D model of the world, not just GPS.

The Trick That Makes Google's Self-Driving Cars Work
By Alexis C. Madrigal, May 2014, The Atlantic
"Rather than having to figure out what the world looks like and what it means from scratch every time we turn on the software, we tell it what the world is expected to look like when it is empty," Chatham continued. "And then the job of the software is to figure out how the world is different from that expectation. This makes the problem a lot simpler."

...


Like what google is doing is inefficient to begin with, but it would make automated city driving functional. Look for a future where you can add a 200K automated driver to your luxury car, but it will only work in the city that you bought the car in.

[deasine]

The discussion about automated vehicles ends here. Back on topic please.

[ninjakafi_81]

Any word on the brand new buses we're getting sometime end of this year or early next ?

[twoNeurons]

Originally Posted by aberdeen5698

This is what I just don't get about autonomous vehicle advocates - they seem to have these very high expectations of space savings due to automation, yet they're ignoring all of the safety and reliability issues that this implies. Automatic train control in protected rights of way is a mature technology, but even the most aggressive systems rely on a "moving block" exclusive occupancy that guarantees enough distance for a following vehicle to come to a safe stop in the event of a failure.

Yet we're somehow expected to believe that autonomous vehicles in a much harsher threat environment are going to magically be able to zip around with merely a few feet of separation and whiz through intersections in an interleaved pattern at full speed? That the propulsion, braking and sensor systems these privately owned vehicles will rely on to operate safely will never fail? That these will all operate at 100% efficiency in inclement and slippery weather?

I can envisage this sort of thing happening some day, but nowhere near as quickly as a lot of people seem to believe.

VERY WELL SAID.

Not to mention failsafes in the case of sabotage, the ability to manually override controls, the margin of error for a myriad number of systems.

[Tfreder]

To get back on topic, the 2013 Translink System Performance numbers have been released.

So far, the 555 only has 565,000 riders per year. The only route I can really think to compare it to is the 351, which had 1,463,000 riders last year. EDIT: considering the 555 receives 12,300 revenue hours per year, while the 351 receives 50,200, I guess the 555 is doing pretty well.

Based off of ridership data from only half of the year (Sept-Dec), Translink estimates that the annual ridership of the 96 B-line is hovering around 2,889,000. Not bad, I suppose.

[xd_1771]

^ I think by that measure the 555 is a pretty big success.
You really have to consider that the 351 has the ability to attract direct bus-only rides because it extends local service into South Surrey down King George Blvd, 152nd St and 16th Ave - whereas with the 555 you must either walk, drive or bus to the park-n-ride.

[twoNeurons]

Considering it's a brand new route, it's very successful:

Cost & Performance

2013 ( Rank )
AnnualBoardings: 565,000 (96 of 206)
Cost/BoardedPassenger: $2.13 (106 of 206)
Avg.DailyBoardings(Mon - Fri): 2,000 (94 of 204)
Avg.DailyBoardings(Saturdays): 700 (105 of 172)
Avg.DailyBoardings(Sun&Holidays): 500 (106 of 163)
PeakPassengerLoad(BiͲdirectionalAvg.): 19 (108 of 204)
PeakLoadFactor(VehicleOccupancy): 40% (104 of 204)
AverageCapacityUtilization(PassengerTurnover): 40% (137 of 204)

Essentially, it's hovering around the 50th percentile for most statistics.