I was reading an article just the other day about Google and other companies that are trying to build an autonomous 'smart' car that can drive itself, and how such cars have already been prototyped out to some success in some areas along the west coast. The article was talking about such computer-controlled cars being commonplace in probably about 20 years. As fast as technology evolves, that wouldn't surprise me at all.

This got me to thinking about the arguments for mass transit... and in a lot of ways, the autonomous car is probably the auto-manufacturers response to a lot of them. It doesn't address the benefits to city infrastructure and things of that nature that mass transit bring, but from an end user standpoint a lot of the boxes are ticked... so the question is, what kind of impact will this type of smart car have on mass transit? Is it a rail killer?

Discuss....

I doubt it would make it much worse for mass transit than the current situation, LA proves that people are willing to drive ridiculously long commutes through congested roads, depending on how implemented it should increase the capacity of all road but most cities/states have not backed off widening roads so it would have only shifted how the capacity was added.

It could also get people use to having less direct control of their transportation leading to more acceptance of mass transit. The push for mass transit and denser cities is already showing support in many cities even with existing federal dollars for roads and utility projects, the high probability that those monies are either shrinking or going away only makes it more pressing for cities to adjust how they spend money going forward.

the big point is with autonomous cars that talk to eachother .. traffic speeds could increase commute times would drop and we would make a much more efficient use of the roads we have ..

On days when I depart at 6-ish because I need to be a few counties away by 8:30ish, a car that drove itself would let me depart quite a bit later. The ability to snarf something down and read in route safely would be rather welcome.

Initial reaction: Yes ... but after reading the above comments, it depends on who owns the autonomous cars. They could BE the future of mass transit. No need to plan routes. Simply have them respond like taxis. You could log in a request (on your smart phone?) and within 15 minutes, a car is there for you. Reserve one in advance noting your embarkation point, destination and requested time window and the computer could match it to similar ones, send back a message as to how it (the mass transit authority's autonomous car system) will satisfy your request and it could send an autonomous sedan, van or bus to meet multiple similar demands. The fee could be less with shared rides or a fee could be assessed if you failed to use an executed request.

The ability to snarf something down and read in route safely would be rather welcome.

This IS mass transit in cities where it functions properly.

Why not just do skytran and forget the car.

SkyTran – Futuristic Public Transport with Maglev System | Tuvie (http://www.tuvie.com/skytran-futuristic-public-transport-with-maglev-system/)

Initial reaction: Yes ... but after reading the above comments, it depends on who owns the autonomous cars. They could BE the future of mass transit. No need to plan routes. Simply have them respond like taxis. You could log in a request (on your smart phone?) and within 15 minutes, a car is there for you. Reserve one in advance noting your embarkation point, destination and requested time window and the computer could match it to similar ones, send back a message as to how it (the mass transit authority's autonomous car system) will satisfy your request and it could send an autonomous sedan, van or bus to meet multiple similar demands. The fee could be less with shared rides or a fee could be assessed if you failed to use an executed request.

That is the perfect transportation system Dubya. It essentially takes the manufacturing engineering principle of the "lot of one" and applies it to transportation. But if people think present day transit systems are expensive, imagine the technology, infrastructure, and potential liability costs of that one. It would be a very interesting engineering exercise to research the viability of a system such as that though.

Oops - double post.

Why not just do skytran and forget the car.

SkyTran – Futuristic Public Transport with Maglev System | Tuvie (http://www.tuvie.com/skytran-futuristic-public-transport-with-maglev-system/)

Cool, but with autonomous cars, I don't have to travel to get to a terminal.

Honestly - I don't see autonomous cars working because you still have to share the road with non-autonomous cars and how much would it suck to be stuck in a car about to get hit by a texter and there is nothing you can do about it except brace for impact. Plus, unless these cars are going to drive on something beside 4 tires and a road, it doesn't solve any of the wear and tear issue on our nation's infrastructure and I assume they would all still need to run on gasoline. Finally, what problem is actully be solved by one of these cars - just the physical driving part? That is the only part people enjoy. Let me also add, I work in a building with 4,000 people - are 4,000 autonomous cars going to line up everyday to take us all home?

... are 4,000 autonomous cars going to line up everyday to take us all home?

Nope. They'll just text their owners about what happy hour parking lot they'll be waiting in.

Bored, looking through old posts and saw this.

Theoretically your car could take you to work and then go back home, eliminating the need for a parking space downtown. If you live too far away, it could drive itself to some big mega-garage in the middle of nowhere. You text your car when you're ready for it to come pick you up, and it follows GPS instructions from your phone and comes to get you. It doesn't eliminate the problems of traffic, in fact it probably increases them. You'd extend rush hour as you have cars driving around by themselves long after the people who used them were already at work.

Personally I just don't see them as feasible anyway. What happens when a sensor goes out? You'd have a car flying down the road oblivious to whatever is in front of/beside it. They're supposed to communicate with each other, but what happens if a deer runs across the road? What happens if someone gets in your car while it's sitting at a red light?

... You'd have a car flying down the road oblivious to whatever is in front of/beside it. ...

Not sure this part is all that different from today for some folk.
:)

What if the autonomous car hits a child, gets a flat, or is involved in an accident?

What if the autonomous car hits a child, gets a flat, or is involved in an accident?

It will likely have lower statistics of hitting children than human drivers and a better chance of reacting before it happens; not that it maters after the fact for the family who did it.

Getting a flat is much easier to test and have test responses for than other things that may arise. Plus since it will almost certainly have restraints on how it drives is less likely to get flats we in the worst situations that may occur.

Until the legalities of who is responsible for damages is settled they will not be sold.

Cool, but with autonomous cars, I don't have to travel to get to a terminal.

Yes you do, you just aren't driving. You will still wait in the same traffic. The only difference is you will be 3 feet to the right.

It will likely have lower statistics of hitting children than human drivers and a better chance of reacting before it happens; not that it maters after the fact for the family who did it.

I have been hit by two cars while riding my bike. In both incidents the drivers didn't have enough time to react and avoid me, but both times they stopped and helped me. Is an autonomous car going to stop and help me? Will it even know it hit me or would I have been chalked up to a bump in the road?

I have been hit by two cars while riding my bike. In both incidents the drivers didn't have enough time to react and avoid me, but both times they stopped and helped me. Is an autonomous car going to stop and help me? Will it even know it hit me or would I have been chalked up to a bump in the road?

You're going to have to get an autonomous bike.

You're going to have to get an autonomous bike.

That is what led to my crashes in the first place. No one was driving the bike :)

chalked up to a bump in the road, or a chalk outline of a bump in the road?

I have been hit by two cars while riding my bike. In both incidents the drivers didn't have enough time to react and avoid me, but both times they stopped and helped me. Is an autonomous car going to stop and help me? Will it even know it hit me or would I have been chalked up to a bump in the road?

An estimate for where they were five years ago to stop was from initial detecting an object was 1/6th the time the best possible human reaction could be, that might even be improved upon, also their is a good chance it would notice you sooner than a person would as well and less likely to get itself in a position to hit you. It will likely be programmed to stop if a collision is either detected or radar/FLIR indicated one might have been possible, the person may still drive off.

chalked up to a bump in the road, or a chalk outline of a bump in the road?

Well technically, I ended up on the sidewalk both times.

An estimate for where they were five years ago to stop was from initial detecting an object was 1/6th the time the best possible human reaction could be, that might even be improved upon, also their is a good chance it would notice you sooner than a person would as well and less likely to get itself in a position to hit you. It will likely be programmed to stop if a collision is either detected or radar/FLIR indicated one might have been possible, the person may still drive off.

Applying brakes and stopping in time are two different things. A car doing 45 mph takes about 100 feet to stop. If I am in that 100 feet no amount of computer programming is going to help.

But anyhow, back to the originial premis - how are automomous cars going to reduce traffic and kill mass transit. Like I said earlier, the office complex I work in has 4,000 people. Are 4,000 autonomous cars going to pick us up? How is that different than the 4,000 cars we already drive?

I have been hit by two cars while riding my bike. In both incidents the drivers didn't have enough time to react and avoid me, but both times they stopped and helped me. Is an autonomous car going to stop and help me? Will it even know it hit me or would I have been chalked up to a bump in the road?In Oklahoma?

I have absolutely no interest in them, but then I am a "car guy" and prefer to be in control of my vehicle. I can definitely see how that type of car would be better than many of the so-called "drivers" currently would be.

I have been hit by two cars while riding my bike. In both incidents the drivers didn't have enough time to react and avoid me, but both times they stopped and helped me. Is an autonomous car going to stop and help me? Will it even know it hit me or would I have been chalked up to a bump in the road?
In Oklahoma?


No, both times were in California.

I like this idea of an autonomous cars by google. However, I have to disagree on where some of you are taking this idea. This is just my opinion, but I would never want to give up my personal vehicle. Having to rely on a system/technology to get me where I want to be bothers me. I love my independence that comes with owning a vehicle. Now, if we can set up vehicles that can switch between automatic and manual driving (computer vs human), that would be the ultimate solution in my mind. I would like it if we could switch to automatic driving for the highway and we can control our vehicle for city streets or something. That would help tremendously with commute times and efficiency. Then, if you could add in a "smart" traffic light system for city streets on top of that, I think our commutes would be tremendously better.

I just personally do not like the idea of giving up my personal vehicle.

I like this idea of an autonomous cars by google. However, I have to disagree on where some of you are taking this idea. This is just my opinion, but I would never want to give up my personal vehicle. Having to rely on a system/technology to get me where I want to be bothers me. I love my independence that comes with owning a vehicle. Now, if we can set up vehicles that can switch between automatic and manual driving (computer vs human), that would be the ultimate solution in my mind. I would like it if we could switch to automatic driving for the highway and we can control our vehicle for city streets or something. That would help tremendously with commute times and efficiency. Then, if you could add in a "smart" traffic light system for city streets on top of that, I think our commutes would be tremendously better.

I just personally do not like the idea of giving up my personal vehicle.

I get you, but just because you have a car doesn't mean that you have to use it. In my little McWorld, I would like to have a bike, econo-commuter car, family car, and a work truck. I see a need for all of them, but that doesn't mean that I wouldn't love to use an on-demand autonomous car for transit some -- maybe even a lot. It's just a matter of having the right tool at my disposal. I'll bet that, in my little McWorld (hey, it could happen), I would eventually not want the two cars, but I wouldn't want to do without them tomorrow. What if I could demand an autonomous family car on demand, too? What if there were a truck-share or tractor-share program near to my land. Then I might not "need" any of them.

Herman Cain's "Job Creators Network" has a somewhat conservative take on the driverless car making mass transit obsolete.
Google?s RoboCar: The Meteor that Could Wipe Out Mass Transit | Job Creators NetworkJob Creators Network (http://jobcreatorsnetwork.com/googles-robocar-the-meteor-that-could-wipe-out-mass-transit/)

Well, unless these autonomous cars run on lollipop wishes we aren't going to be able to afford to have the drive themselves because we can't keep funding the US military at the rate we have been. I wish my fellow Tea Party friends would wake the hell up and realize the ‘private automobile’ is one of the largest expenses of the federal government.

First time I've seen this thread...I like, I like.

First thought that comes to mind. It won't work until everything is 100% autonomous from coast to coast. Either that or you build the infrastructure in sections and then they say, "autonomous cars can only operate in these sections".

Secondly, if converting only certain sections over one at a time or in regions and expanding outward to connect all sections, the first autonomous cars would cost more than any car on the market right now. I only say this because that car is going to have to have a supercomputer on it that has a base CPU of so many teraflops to handle the operations of making accurate calculations to avoid other autonomous cars and other non autonomous objects (kids, wildlife, pets lost on roads, etc.) and all that sensory equipment placed on every angle of these cars is going to cost a lot of money.

Thirdly, if this thing is going to be operated by a network, then it is going to have to be super hacker protected cause this would be a major terrorist target.

I don't think it would kill rail because there will be people that couldn't afford the autonomous car and they would choose to use other mass transport options like rail and bus.

I honestly think an 'I, Robot' transportation world is still 100+ years in the future. I remember when I was a Boy Scout and Boys Life magazine had an issue where the cover of it had a futuristic autonomous car. At least they were able to dream in the 1980s.

The autonomous A.I. will have to advance a bit more. Right now its hard to beat a healthy, trained human mind. At least for reaction time and options sorting ability. It will probably be awhile.

First time I've seen this thread...I like, I like.

First thought that comes to mind. It won't work until everything is 100% autonomous from coast to coast. Either that or you build the infrastructure in sections and then they say, "autonomous cars can only operate in these sections".

Secondly, if converting only certain sections over one at a time or in regions and expanding outward to connect all sections, the first autonomous cars would cost more than any car on the market right now. I only say this because that car is going to have to have a supercomputer on it that has a base CPU of so many teraflops to handle the operations of making accurate calculations to avoid other autonomous cars and other non autonomous objects (kids, wildlife, pets lost on roads, etc.) and all that sensory equipment placed on every angle of these cars is going to cost a lot of money.

Thirdly, if this thing is going to be operated by a network, then it is going to have to be super hacker protected cause this would be a major terrorist target.

I don't think it would kill rail because there will be people that couldn't afford the autonomous car and they would choose to use other mass transport options like rail and bus.

I honestly think an 'I, Robot' transportation world is still 100+ years in the future. I remember when I was a Boy Scout and Boys Life magazine had an issue where the cover of it had a futuristic autonomous car. At least they were able to dream in the 1980s.

A group I heard about was working on it in the 90's indicated that it will likely be in phases like first the interstates and sections of other highways where there are more limits to what should be there, is pretty standardized across the country and takes less time to affect large numbers of people. With city streets likely being the last thing to happen, possibly many years after the others.

Some standard methods will eventually need to be chosen, there have been several methods done by various groups from using strong permanent magnets in the street like the white dotted lines for precision guidance to pure camera systems that just read existing lane markers. Several already work pretty well with less computer power than an average PC, granted some of those were using closed courses. In the end, I doubt computer power is as much of an issue as getting all the questions about legal liabilities settled.

One thing that DOTs have been interested in is how they can communicate with each other to give advance warning of needs to slow or stop to pair with them already having better reaction time than humans to give lanes that are totally automated a much higher volume per lane than humans could drive. Which has the downside if that ever completely messed then best case you then will have a road that will likely be nowhere close to the capacity it would have after a several years of use and at worst a risk of a massive pile up.


The autonomous A.I. will have to advance a bit more. Right now its hard to beat a healthy, trained human mind. At least for reaction time and options sorting ability. It will probably be awhile.

Unfortunately I have some doubts on autonomous AI being behind a lot of the driver I see on the road, it is hard to drive more than few miles consecutively anywhere with traffic before you find someone driving distracted, impaired, tailgating, failing to yield, bizarrely slow reactions to any intersection/turn/time to accelerate, etc.