Should the coming self driving revolution cause us to rethink transit plans?

[fuji]

LOL...is this a joke??

What do you need a drivers license for if you're not actually driving the car??

Presumably they licence somebody to be the operator. It's an interesting question how licensing will work with self driving cars. But the story is true, the province is allowing self driving cars to be tested by the university.

As for those who say they are nostalgic for stick shifts and the manual driving experience, they can always go to a race track. I understand you can also find places to ride a horse and I bet there are even places that let you try your hand at driving a horse and buggy.

Sounds like a fun family outing in 2037, grandpa takes the grand kids out to a track so they can experience old fashioned cars and they look on in wonder as he talks about steering wheels, brakes, even gear shifts all being controlled by the driver in the old days.

 

[Brill]

Self driving cars can be a lot smaller, why not a 4 ft x 4 ft if it's only one person? They could be coupled or stacked, you might fit 20 of them in the area of a current car.
For those who say they enjoy driving, I'm guessing they don't commute in Toronto rush hour.

 

[Butler1000]

That happens to human drivers too. But self driving cars have a huge advantage: if they are all connected and being managed by the system they can all brake at the same time so that they don't pile up.

And they can brake in response to the onboard diagnostics on the failing vehicle. Meaning within a millisecond of the lead vehicle detecting loss of traction or loss of tire pressure all the vehicles will receive a signal to slow down and they will do so in unison. As if the brakes in your car were being controlled by the brake pedal in the car six vehicles ahead.

Moreover the presence of ice will be remembered by the system and all the vehicles will adjust accordingly for the rest of the day at that spot.

Connected cars and a smart roadway simply have a lot more information than you do and can use it faster and more intelligently.

As for long distances that's likely to be the first problem solved as the trucking industry has a huge financial interest in automated long haul trucking.

You are talking about speeds of 200k/m. Ever hear of inertia? Cars can't stop that fast period.

It's one of the reasons we have speed limits.

It's a nice thought. But 20 years? Doubtful. It isn't about technology. It's about changing people. And while there maybe enough inertia to allow for limited quantities in urban environments you can forget rural and suburban.

 

[Butler1000]

Seems the province just issued the first driverless drivers licenses to u of waterloo.

Three vehicles to be tested by three different companies(including blackberry, could their encryption tech be of use here? May be time to think long game on the stock).

It's a TEN year pilot project to test them and see how they perform.

Like I said. A long way off.

 

[fuji]

You are talking about speeds of 200k/m. Ever hear of inertia? Cars can't stop that fast period.
.

Which is fine because the car having problems is subject to inertia as well and also won't stop that fast. If it slams on its brakes yours will slam on at the same time. If you start braking just it loses control you will stop before it does.

That means you can either significantly reduce the following distance OR significantly increase speeds because reaction time is going to be at least a thousand times faster (milliseconds instead of seconds).

 

[GameBoy27]

Toronto is setting out to spend hundreds of millions on new transit lines: light rail and/or subway lines. We have had the light rail vs subway debate many times, but are they both wrong?

Are we building transit for the 20th century instead of the 21st?

Self driving vehicles seem certain to revolutionize transportation. They aren't just a little bit different than what we have today. They are a LOT different.

They will result in a totally different way of thinking about transportation just as these hugely expensive projects come fully online--twenty years from now.

Twenty years is a long time. Technology is moving fast.

Self driving cars will change the way we think about car ownership. It's going to become a leasing model for a lot of people, like a ZipCar that comes to you when and where you want it. Like Uber, but without a driver.

Drivers are the big operating cost in public transit, whether we are talking about buses or taxis, It's the cost of the driver that is usually the limiting factor in how many are on the road.

Or flip it around. Assuming ALL cars can self drive, why wouldn't you send YOUR car out to make some money while you aren't using it? YOUR car spends most of its existence unproductively parked. Wouldn't it be better if it could go drive around for Uber for a few hours earning some fares while you are at work?

Instead of paying for parking, you'd get paid.

So what are the implications of that:

1. No more street parking. Nobody will ever park. Their car will drop them off and drive away, maybe to pick up other people. Maybe to park somewhere else.

2. Ride sharing replaces GO train, subway, and buses. Why take any of those things when you can get door to door service? A self driving minivan can pick up and drop off a few people on a common route to work and do so cheaply.

Maybe we will build a hugely expensive subway system and no one will ever take it because it will be obsolete before it's even built.

Maybe instead we should begin investing in self driving infrastructure. Traffic controls that can talk to a computerized car, upgrades to the highways to make them easier for computer driven cars to navigate, off street parking designed for self driving cars to go stow themselves until needed--Wi-Fi enabled so they can receive your call to come pick you up.

By eliminating all street parking we gain a lot of capacity on the roads. Maybe we don't need trains.

Dear God fuji. Can I have some of what you're smoking? While I agree that self-driving cars, if they come with an option to turn it on or off, my guess is that a lot of people will be turning it off. Drivers are impatient at the best of times. I can imagine many drivers would get tired of the vehicles never exceeding the speed limit, never accelerating too quickly, never changing lanes to get around a slower moving vehicle, drving up on the sidewalk to get around a garbage truck on a narrow street etc.

In addition, they're non-aggressive which you somewhat need to be when trying to make a right turn, with a constant flow of pedestrians on downtown streets.

Your idea of moving people faster by eliminating subway cars and replacing them with self-driving vehicles is so far off the mark. A subway train holds up to 1500 people and can run every 90 seconds. You would need the equivalent of 250 vehicles carrying 6 people in each to match 1 subway train. It makes no sense.

Then there's the affordability aspect. Self driving cars will always be considerably more expensive than bare bones economy car. So you'll always have a mix of regular and self-driving cars. I'm not saying there won't be self-driving cars, just that it's not going to be as prevalent as everyone thinks.

And I still wonder how well self-driving cars will work in conditions like this. Or when cars get covered with snow, ice or splattered with thick amounts of dark brown slush from passing cars while driving one the highway?

 

[huckfinn]

Which is fine because the car having problems is subject to inertia as well and also won't stop that fast. If it slams on its brakes yours will slam on at the same time. If you start braking just it loses control you will stop before it does.

That means you can either significantly reduce the following distance OR significantly increase speeds because reaction time is going to be at least a thousand times faster (milliseconds instead of seconds).

Problem is, in order for this to work properly, all the cars would have to have the same stopping distance / capability regardless if they are light or heavy.

Also, the condition of all their brake pads, calipers, tires etc would have to maintained exactly the same to get the same braking power.

Lots of variables.

 

[explorerzip]

Dear God fuji. Can I have some of what you're smoking? While I agree that self-driving cars, if they come with an option to turn it on or off, my guess is that a lot of people will be turning it off. Drivers are impatient at the best of times. I can imagine many drivers would get tired of the vehicles never exceeding the speed limit, never accelerating too quickly, never changing lanes to get around a slower moving vehicle, drving up on the sidewalk to get around a garbage truck on a narrow street etc.

In addition, they're non-aggressive which you somewhat need to be when trying to make a right turn, with a constant flow of pedestrians on downtown streets.

Your idea of moving people faster by eliminating subway cars and replacing them with self-driving vehicles is so far off the mark. A subway train holds up to 1500 people and can run every 90 seconds. You would need the equivalent of 250 vehicles carrying 6 people in each to match 1 subway train. It makes no sense.

Then there's the affordability aspect. Self driving cars will always be considerably more expensive than bare bones economy car. So you'll always have a mix of regular and self-driving cars. I'm not saying there won't be self-driving cars, just that it's not going to be as prevalent as everyone thinks.

And I still wonder how well self-driving cars will work in conditions like this. Or when cars get covered with snow, ice or splattered with thick amounts of dark brown slush from passing cars while driving one the highway?

Fuji's point is that people would not be purchasing self driving cars. People would either be subscribing to a monthly / yearly service or there could be a pay per ride option. There are times when I honestly would be interested in such a service because I would no longer have to worry about changing tires, oil or other maintenance. It's just a pain to have to do those things. And I could let the car do all the work while I take a nap after a long day at work, or do other things with my time.

The ideal situation is that people won't be able to change the driving dynamics of a self-driving car to pass a slow moving vehicle, pedestrians, etc. You are right that if given a choice, some people will disable the automatic driving tech.

The solution to bad weather is not that difficult to solve because we already have technologies like TCAS for aerospace and AIS for marine. Both are basically vessel identification and collision avoidance systems. So I'm sure you could program a self-driving car to maintain a specific amount of distance between other vehicles regardless of weather conditions. People tailgate and drive aggressively. Computers can't do that.

 

[fuji]

Your idea of moving people faster by eliminating subway cars and replacing them with self-driving vehicles is so far off the mark. A subway train holds up to 1500 people and can run every 90 seconds. You would need the equivalent of 250 vehicles carrying 6 people in each to match 1 subway train.

Let's do the math. It takes 45 minutes for a subway to travel the 37km from Islington to Kipling. During that 2700 minutes a total of 30 trains will arrive carrying a total of 45,000 people.

So how many minivans do you need to move 45000 people in 45 min assuming that they travel at 150kmph non stop the whole way (entering and exiting vehicles don't stop but instead merge at highway speed)?

We need them to arrive at a rate of 2.778 per second, or one minivan every .36 seconds. A minivan traveling 150kmph is moving at a rate of 41.7 meters per second, or roughly 15 meters in the .36 seconds between arriving minivans.

Assuming a minivan is about 5m long and we need one minivan per 15m there will be 10m between minivans, or roughly two car lengths. Seems doable in a system controlled by self driving systems and managed by a smart roadway.

Are we there yet with the technology we have today? Almost. My car can already do this with adaptive cruise control. Add a smart highway to assist with merging and traffic management and you are in business.

And the self driving system has numerous other benefits:

1. Non stop travel, every vehicle is an express, you don't stop for others to get on or off

2. No changing from a bus to a subway, you take one vehicle the entire way

3. No walking to your stop, it picks you up and drops you off at your door

4. Faster travel times, a minivan can go 150kmph, while a subway simply can't go that fast

 

[fuji]

Problem is, in order for this to work properly, all the cars would have to have the same stopping distance / capability regardless if they are light or heavy.

Also, the condition of all their brake pads, calipers, tires etc would have to maintained exactly the same to get the same braking power.

Lots of variables.

Computers are good at managing lots of variables and can easily allocate more stopping distance in front of a truck than in front of a smart car.

I agree with the maintenance point which is why the accelerated lane is likely going to be restricted to vehicles that are part of a commercial fleet so that they receive this maintenance. If you want to travel in that fast lane you need to summon a TTC vehicle or one that is part of an approved and managed taxi fleet.

 

[fuji]

Fuji's point is that people would not be purchasing self driving cars.

Some people might still have reason to do that. For example a contractor who carries all his own tools in his truck would want to at least use the same truck day after day and therefore likely own or lease it.

These private vehicles might not be allowed on some of the optimized roadways and would instead stay on the regular roads along with human driven vehicles. Driver might sometimes take over (e.g., driving into the construction work site or into farm land) and at other times allow the vehicle to take over (cruising on the highway).

But many people would summon vehicles on demand and even the owner of the truck would prefer that solution for a Friday night out drinking with his buddies.

 

[Phil C. McNasty]

Question is, would anyone wanna be driving a missile (thats your car) at 120 km/h on the 401 knowing the only thing thats separating you from death is a computer bug??

 

[Butler1000]

Fuji, they just authorized a ten year study. Which means they won't allow them on the roads at all for ten years. Then you have to ramp up production and the natural attrition of cars. Then convince people they are safe and worth it.

Then you have to convince the politicians to change numerous laws guaranteed to piss people off, not to mention accept the risk of political fallout if failure occurs and people get hurt or worse.

It won't happen in 20 years.

 

[GameBoy27]

Let's do the math. It takes 45 minutes for a subway to travel the 37km from Islington to Kipling. During that 2700 minutes a total of 30 trains will arrive carrying a total of 45,000 people.

So how many minivans do you need to move 45000 people in 45 min assuming that they travel at 150kmph non stop the whole way (entering and exiting vehicles don't stop but instead merge at highway speed)?

Let's do the reality. It's just over 26 kms (not 37) between Kipling and Kennedy stations. I believe that's what you meant, although I'm not sure what 2700 minutes has to do with anything. Anyway, the point of subway stations is to allow people to get on or off where they please. I think you'd find there's only a small percentage of total riders who commute the entire distance. Your logic is failed.

Fuji, they just authorized a ten year study. Which means they won't allow them on the roads at all for ten years. Then you have to ramp up production and the natural attrition of cars. Then convince people they are safe and worth it.

Then you have to convince the politicians to change numerous laws guaranteed to piss people off, not to mention accept the risk of political fallout if failure occurs and people get hurt or worse.

It won't happen in 20 years.

100% ^^^

I applaud Toyota for installing automatic emergency braking (AEB) as a standard feature on almost all of its cars by 2017. That's 5 years ahead of automobile manufacturer's agreed time frame. Since 25%-30% of all collisions (which I think is a conservative estimate) are rear-ends, this is great news!

http://www.theglobeandmail.com/globe-drive/culture/technology/base-model-toyotas-to-come-standard-with-full-suite-of-safety-features-by-2017/article29864419/

 

[Phil C. McNasty]

Fuji, they just authorized a ten year study. Which means they won't allow them on the roads at all for ten years. Then you have to ramp up production and the natural attrition of cars. Then convince people they are safe and worth it.

Then you have to convince the politicians to change numerous laws guaranteed to piss people off, not to mention accept the risk of political fallout if failure occurs and people get hurt or worse.

It won't happen in 20 years

Fuji is wrong??? That cant be, he's always right

I applaud Toyota for installing automatic emergency braking (AEB) as a standard feature on almost all of its cars by 2017. That's 5 years ahead of automobile manufacturer's agreed time frame. Since 25%-30% of all collisions (which I think is a conservative estimate) are rear-ends, this is great news!

http://www.theglobeandmail.com/globe-drive/culture/technology/base-model-toyotas-to-come-standard-with-full-suite-of-safety-features-by-2017/article29864419/

If a picture is worth a thousand words, then a video is worth a million

 

[SirWanker]

So how will this driverless car concept deal with:

• cyclists
• motorcycles
• delivery vehicles

Better yet how will one deal with extremely busy drop off zones e.g. sporting events, concerts or situation like Union Station in rush hours?

 

[fuji]

Question is, would anyone wanna be driving a missile (thats your car) at 120 km/h on the 401 knowing the only thing thats separating you from death is a computer bug??

I do that every time I get into an airplane except it's 900km/h. No problem.

 

[GameBoy27]

So how will this driverless car concept deal with:

• cyclists
• motorcycles
• delivery vehicles

Better yet how will one deal with extremely busy drop off zones e.g. sporting events, concerts or situation like Union Station in rush hours?

All good points ^^^

And:

- know to pull over and stop for a RIDE spot check
- know to keep a 1m distance from cyclists
- know to move over on multi-lane highways when approaching/passing emergency vehicles
- know to treat an intersection with traffic lights (if they change to flashing red) as a four way stop
- know to drive through a red light when a police officer on point duty instructs you to
- know to obey no turn signs on specific days and times which can be changed without notice
- know to obey temporary no turn signs are placed in construction zones
- know to obey speed limits on roads when they they change either permanently or during construction
- know to drive in the oncoming lane when overnight construction requires a detour
- decide to either panic stop/swerve (potentially causing a pileup behind) or run over an item that fell off the back of a truck, or an animal crossing the road

Those are just a few of the hurdles they'll have to overcome.

 

[fuji]

Let's do the reality. It's just over 26 kms (not 37) between Kipling and Kennedy stations. I believe that's what you meant, although I'm not sure what 2700 minutes has to do with anything.

Typo, sorry, 2700 seconds in 45 minutes.

What I was calculating was how frequently cars would have to arrive in order to deliver as many people as a subway. It turns out that you need one car to arrive every .36 seconds to deliver the 45000 people a subway can deliver in 45min.

Then I calculated how far apart those cars would be to see if one lane had the capacity to deliver cars that fast. It turns out that at 150kmph you will have one 5m car in every 15m interval meaning cars will have two car lengths between them.

That's totally doable.

Anyway, the point of subway stations is to allow people to get on or off where they please. I think you'd find there's only a small percentage of total riders who commute the entire distance.

The self driving system is even better. Your car picks you up at your door, drives to the subway, enters on the cloverleaf that replaces the subway station, merges with the existing traffic at highway speed, drives at highway speeds through all stations to your exit (merging with other vehicles entering and exiting), then exits at a cloverleaf, drives onto the city steet, and takes you right to the door at your destination.

Basically turns the subway into a highway restricted to self driving vehicles that are integrated with one another and with the tunnel itself so that they can safely travel at 150kmph spaced two car lengths apart. Instead of a subway signaling system managing trains you have a tunnel system managing the entry and exit of cars.

Much faster than a subway, and much more convenient.

 

[Aardvark154]

Lord isn't this George Jetson pie in the sky.