Driverless Car Poll Finds Canadians Almost Evenly Split On Issue

[Butler1000]

In Toronto/Gta we could use something like high speed trains, conjestion is an issue.

Um, have you seen how many people already get hit by GO trains?

Anyway the tech involved is expensive. And like I said the density isn't there. The purpose is to transport from at least X number of miles away enough people to a city centre as a commute to make it financially viable. It doesn't work within a city because of constant stop and go pick up and drop off.

So unless the pop of Newmarket, Barrie et al grows to a Million and needs to come to Toronto to Work (like Japan with Tokyo) it ain't worth it.

 

[fuji]

I already have that....

It's called the TTC.

No you don't, it doesn't come get you wherever you are and drive you across the city to where you want to go. You catch an irregular bus to a mainline bus or subway to another buyer. Multiple transfers.

A self driving car service would pick you up near your home and drive you to near your destination. In future once all roads are built it would go door to door. And it would go door to door if your end points are on mapped streets.

TTC might wind up being the operator and the self driving cars would potentially replace the feeder buses that her people to the mainlines.

 

[fuji]

Again, what is the advantage over existing transportation options?

Cost. No driver. Cheap. Your can also carry an extra passenger since space isn't wasted on a driver.

In fact self driving buses are likely to be one of the first applications since the route is predictable and can be instrumented with electronic signals to aid the self driving bus, solving many of the initial problems. Human drivers would run on the routes with any difficulties (eg construction that day).

According to everything I've read so far, determining the difference between a police officer, a construction worker, a crossing guard and a homeless guy is proving to be quite a challenge.

Give the police electronic badges that the car can recognize. Ditto construction crews on the highway who want to stop cars.

These are problems to be solved but none of them are hard to solve.

You would also be surprised how good recognition technology is getting. Not only can software recognize a person, it is getting pretty good at facial recognition and identifying WHICH person. Not only will the car be able to recognize an officer, if given access to a database, it could tell you his name.

Recognizing an officer if bit hard but giving him an electronic badge makes it bullet proof.

 

[Butler1000]

Cost. No driver. Cheap. Your can also carry an extra passenger since space isn't wasted on a driver.

In fact self driving buses are likely to be one of the first applications since the route is predictable and can be instrumented with electronic signals to aid the self driving bus, solving many of the initial problems. Human drivers would run on the routes with any difficulties (eg construction that day).



Give the police electronic badges that the car can recognize. Ditto construction crews on the highway who want to stop cars.

These are problems to be solved but none of them are hard to solve.

You would also be surprised how good recognition technology is getting. Not only can software recognize a person, it is getting pretty good at facial recognition and identifying WHICH person. Not only will the car be able to recognize an officer, if given access to a database, it could tell you his name.

Recognizing an officer if bit hard but giving him an electronic badge makes it bullet proof.

I don't think you realize the value of the driver beyond driving.

And that's prevention and maintenance.

To whit. The driver is also there to clean up when someone:

Gets sick, makes a mess, their pet makes a mess etc.

And prevention.....look there is a section of the population who for a lack of better term are filthy animals. They will happily throw up, piss, pick their nose and wipe I on the seat, leave food and spilled drink, masterbate, have sex, leave stickers, grafitti, vandalize.......

And a driver prevents that just with his presence. This is a not a rare thing either. I see this stuff all the time now on the ttc. No driver and you would have to rely on camera monitoring (and someone to do that with the equipment and costs with that), a well as regularly taking the car out of service to clean it for the next customer.

And loss of customers if a cab arrives in a bad state.

Just curious Fuji.....what the hell Is an electronic badge? How does it work, how much does it cost, and who should be responsible for this cost?

And will someone be there to issue them to the public so that during a power outage they can direct them too like what has happened the last couple of times power went out?

 

[Butler1000]

No you don't, it doesn't come get you wherever you are and drive you across the city to where you want to go. You catch an irregular bus to a mainline bus or subway to another buyer. Multiple transfers.

A self driving car service would pick you up near your home and drive you to near your destination. In future once all roads are built it would go door to door. And it would go door to door if your end points are on mapped streets.

TTC might wind up being the operator and the self driving cars would potentially replace the feeder buses that her people to the mainlines.

You stated it would operate on main roads. That isn't door to door. And so it wouldn't initially be better. So most wouldn't bother with it.

It has to start as door to door. Anything les and I will fail.

 

[fuji]

You stated it would operate on main roads. That isn't door to door. And so it wouldn't initially be better. So most wouldn't bother with it.

It has to start as door to door. Anything les and I will fail.

Yes it would be a lot better. A two minute walk to a crowded bus to the crowded subway to another crowded bus, with two transfers, maybe three if you had to change subways, is nowhere near as fast or convenient as a two minute walk to a private car that drives directly to your final stop a two minute walk from your destination.

And that is only a stopgap until it is fully door to door. And if either of your endpoints is on a major street it will be to the door at that end.

 

[fuji]

I don't think you realize the value of the driver beyond driving.

And that's prevention and maintenance.

To whit. The driver is also there to clean up when someone:

Gets sick, makes a mess, their pet makes a mess etc.

And prevention.....look there is a section of the population who for a lack of better term are filthy animals. They will happily throw up, piss, pick their nose and wipe I on the seat, leave food and spilled drink, masterbate, have sex, leave stickers, grafitti, vandalize.......

And a driver prevents that just with his presence. This is a not a rare thing either. I see this stuff all the time now on the ttc. No driver and you would have to rely on camera monitoring (and someone to do that with the equipment and costs with that), a well as regularly taking the car out of service to clean it for the next customer.

And loss of customers if a cab arrives in a bad state.

Just curious Fuji.....what the hell Is an electronic badge? How does it work, how much does it cost, and who should be responsible for this cost?

And will someone be there to issue them to the public so that during a power outage they can direct them too like what has happened the last couple of times power went out?

Cameras and sensors in the car can be used to send it to a garage for cleaning as necessary, billed to the offending passenger as necessary. ZipCar has also solved this just by having customers self report dirty cars and such. Zipcar method does work.

An electronic badge is a transmitter that identifies an officer, using short range radio frequencies. Would likely cost under five bucks and could be built into every officers kit. Similar to the RFID technology in your credit card but with range great enough to be picked up by a car within fifty or a hundred meters, far enough for traffic signaling.

Same technology can be embedded into the street and signs to provide additional electronic traffic controls for computer driven vehicles, and mounted on emergency vehicles to force self driving cars to pull over or stop.

 

[Butler1000]

Yes it would be a lot better. A two minute walk to a crowded bus to the crowded subway to another crowded bus, with two transfers, maybe three if you had to change subways, is nowhere near as fast or convenient as a two minute walk to a private car that drives directly to your final stop a two minute walk from your destination.

And that is only a stopgap until it is fully door to door. And if either of your endpoints is on a major street it will be to the door at that end.

And what's the price point?

 

[fuji]

And what's the price point?

Do the math. Figure that a self driving car would cost five to ten grand more than an equivalent model that isn't self driving. That is about double the 5k price gap between today's driver assist packages and the base model, including the computer and the sensors.

So you have an extra $10k one time capital cost for each taxi cab, but you save what, $30k every year on the driver?

So amortizing the cost of the electronics your operating cost is about $27k lower for the self driving taxi.

Maybe there would be some sort of self driving car tax to pay for those road upgrades, traffic controls and police officers being outfitted with signaling devices. So maybe that is another $2k or 3k per vehicle per year, to pay your share of those upgrades. If Toronto's10k cabs went self driving that would generate 30mil/year to pay to upgrade traffic control.

You still operate 20 to 25k per vehicle cheaper than when you have to pay a driver salary. That is a lot of profit.

 

[Butler1000]

What are you basing the price of the car on? Have you got a link to the hardware cost installed?

And also for the hardware upgrades to the streets. Any links to this? Also how about the data costs incurred? Cost to keep a clean up shop open? Storage? Maintenance?

 

[fuji]

What are you basing the price of the car on? Have you got a link to the hardware cost installed?

And also for the hardware upgrades to the streets. Any links to this? Also how about the data costs incurred? Cost to keep a clean up shop open? Storage? Maintenance?

I took the price of current model computer driving assistance packages available on cars today, that do things like adaptive cruise control, emergency braking, steering adjustment, blindspot monitoring, etc, and doubled it. Obviously we aren't going to have the self driving technology for a number of years but I figure when we do it will cost about the same as the more limited capabilities being sold today. Moore's law.

Data costs are dirt cheap. You can store a detailed map of the earth on a thumb drive these days and buy a good high speed mobile data plan for $50/month.

Taxi companies already operate garages, but sure you would likely have extra staff there doing the cleaning. Maybe one extra detailer for every ten cars on the road?

The overwhelming cost savings and available profits will make this happen. Once it runs as a transit system the technology will then make inroads into other uses as it grows in capability and proves itself.

Private cars will likely be hybrids, before fully self driving, say self driving on highways and major streets but requiring the human driver to take over on side streets and access roads or in any unusual situation. Teslas can aleady automatically perform lane changes, even in heavy traffic, and that kind of capability will grow. Car could pull over or stop at a traffic signal and require the driver to proceed manually when they exit the highway, if the driver's hands and feet weren't already on the controls.

 

[1.8t]

With current insurance regulations there is no way they would allow true driverless transport. someone needs to be accountable. the manufacturers wont do it. so it will require that a driver be at the wheel at all times.
the car will drive it's self but the driver will need to keep an eye on things. maybe even requiring a "dead man switch" on the wheel or pedal that would know if the driver fell asleep. with that in mind, none of the accidents involving cars in test as of now have been the fault of the driverless car.

as for the cop on point duty, I've heard discussion of this in the development groups. the best solution was to alert the car, through a signal from the malfunctioning light or a text to all driverless cars, that the light is out and the car will avoid this area or require the driver take over at that intersection.

all of the other issues, avoiding animals, merging, blocked roads etc are within the scope of the current systems of sensors and processors.

there is no need for any changes to infrastructure as all of the guidance and decision making is aboard the car. it only uses maps the way a current GPS would. position within the road and obstacle avoidance relies on optical and sonar/radar sensors.

 

[Butler1000]

So I've been reading about this Velodyne Laser system the Google car uses. Seems to work. Cost is 75,000+. So until they get get a mass production order from a major car company then I don't see it as viable.

I don't see a mass production order until all 50 states and the Federal Gov't approves their use. So far 2 states, California and Nevada have, but I'm betting it was just for testing.

Add in the 2 radar, camera systems, the onboard computer and the actual cost of the car.......

Then there will be the insurance companies who will be especially reluctant to insure a fleet until they can figure out the pricing model.

It's all very exciting right now but I don't see the inertia coming together on this for awhile.

They did develop a more limited system for the lidar at a reduced cost. This has not been road tested yet. If they can get that one on line then it may have a chance. But not many will get onboard until then.

 

[GameBoy27]

Do the math. Figure that a self driving car would cost five to ten grand more than an equivalent model that isn't self driving. That is about double the 5k price gap between today's driver assist packages and the base model, including the computer and the sensors.

So you have an extra $10k one time capital cost for each taxi cab, but you save what, $30k every year on the driver?

So amortizing the cost of the electronics your operating cost is about $27k lower for the self driving taxi.

Maybe there would be some sort of self driving car tax to pay for those road upgrades, traffic controls and police officers being outfitted with signaling devices. So maybe that is another $2k or 3k per vehicle per year, to pay your share of those upgrades. If Toronto's10k cabs went self driving that would generate 30mil/year to pay to upgrade traffic control.

You still operate 20 to 25k per vehicle cheaper than when you have to pay a driver salary. That is a lot of profit.

I'm doing the math. Since when do taxi drivers make a 30K a year? Other than owner/divers, every other taxi driver rents cabs on a daily basis and earn their wage through the fares they collect. These expensive driverless cars would also have to compete with UberX and regular cabs.

Can't see this as a profitable business venture. Your math is highly flawed.

 

[fuji]

I'm doing the math. Since when do taxi drivers make a 30K a year? Other than owner/divers, every other taxi driver rents cabs on a daily basis and earn their wage through the fares they collect. These expensive driverless cars would also have to compete with Uber and regular cabs.

Can't see this as a profitable business venture. Your math is highly flawed.

30k a year is about right when you look at the total cost of employment. If you would rather assuming 25k though that doesn't change the conclusion. Excess profits with a driverless car would still be in the 20k range.

And renting out the car or hiring it out through Uber won't change the economics. The investor who bought the car makes a lot more money if there is no driver to pay.

 

[IM469]

So I've been reading about this Velodyne Laser system the Google car uses. Seems to work. Cost is 75,000+. So until they get get a mass production order from a major car company then I don't see it as viable. ....Add in the 2 radar, camera systems, the onboard computer and the actual cost of the car.......

Maybe my response should be in the 'you know you're old' thread but 25 years ago a cell phone that had a GPS, 64 bit processor, 32 GB memory, digital camera, Wi-Fi and quad band radio frequency pick up would cost you millions of dollars and be the size of 747. With the market demand pushing it - maybe the intelligent driver could plug in your cigarette lighter and hang from the mirror.

I'm doing the math. Since when do taxi drivers make a 30K a year? Other than owner/divers, every other taxi driver rents cabs on a daily basis and earn their wage through the fares they collect. These expensive driverless cars would also have to compete with UberX and regular cabs.

Can't see this as a profitable business venture. Your math is highly flawed.

I think taxi drivers could follow the path of home milk & bread delivery men. Why would I need them ? If I had my car drive me home from work and I'm staying home, maybe I link it to Uber and the car goes out and makes money for me as I watch Star Trek reruns.

 

[GameBoy27]

30k a year is about right when you look at the total cost of employment. If you would rather assuming 25k though that doesn't change the conclusion. Excess profits with a driverless car would still be in the 20k range.

And renting out the car or hiring it out through Uber won't change the economics. The investor who bought the car makes a lot more money if there is no driver to pay.

Pay what driver? Cab owners don't pay drivers to drive their cars.

 

[fuji]

Pay what driver? Cab owners don't pay drivers to drive their cars.

Is it really your belief that profits to the owner will be the same with and without an additional $25k being subtracted from fare revenue?

 

[GameBoy27]

Is it really your belief that profits to the owner will be the same with and without an additional $25k being subtracted from fare revenue?

You can subtract fare revenues of 25K, but as a plate holder you are losing 35K a year by not renting out your car to drivers at $100 a day. The difference then becomes the price of the car, which will be considerably more expensive for a driverless one.

You also have to factor in that driverless cars will be limited to where they can go, likely for many years. Making them less popular than the current competition being regular cabs and UberX.

 

[fuji]

You can subtract fare revenues of 25K, but as a plate holder you are losing 35K a year by not renting out your car to drivers at $100 a day. The difference then becomes the price of the car, which will be considerably more expensive for a driverless one.

You also have to factor in that driverless cars will be limited to where they can go, likely for many years. Making them less popular than the current competition being regular cabs and UberX.

So, you don't have much of a mind for business. Let me put it this way: when the net cost of operations goes down while revenue remains constant there is more profit. Previously that profit was shared between a driver and an owner. With a self driving car, the owner gets it all. The owner still gets their previous 35k share of the profits, plus now they also get the driver's share, minus the much lower cost of the computer. Their profits are 35+20=55k instead of 35k, using your numbers.

Likely some of the pie would be shared with customers as well, with lower fares, splitting the extra profit between the owners and, in the form of lower prices, the customer. That makes the self driving taxi not only cheaper for customers but also more profitable for owners.

It is true that the self driving car won't initially be able to go as many places. So use an Uber like app where customers call a vehicle entering their destination and pickup address. Dispatch a human driver for the customers who want to go off the grid and a self driving vehicle for the trips it can handle.

You describe Uber, TTC, and taxi companies as alternatives to self driving vehicles, but they will be the operators. Uber in particular is highly likely to be an early adopter, and if they embrace change the TTC and taxi cab companies are natural early adopters as well.