Automating the roads - (split from Flying Cars thread)

[ski_power]

Even I think it would be far safer (both for the general public, and for security purposes) just to automate transport and enclose it. Like said above, if enclosed there is nothing stopping an enclosed system running seamlessly at high speeds.

 

There isn't actually a need for a general flying system of vehicles.The problem that needs to be solved is of safety, and congestion. Both of which automated systems can solve.

 

But then again there isn't need of nuclear warheads either...

[npts2020]

Even a from-helipad-launch would allow entrance into the current private copter market. From driveway would be harder of course, but it is something I am sure many wealthy people would like to have. It may find a market somewhere like Dubai before it does here though.

 

 

 

With massive investment in infrastructure perhaps. It would take me a couple hours by bus to get down town from where I live, or it's a 20 minute drive. As sprawl keeps getting bigger, trying to connect any one place to any one other place gets harder.

 

Mass transit is great for where it works, but the reason people get stuck in rush hour traffic is it's not a perfect fit for all situations.

 

I would like to see improvements in mass transit, it just won't do away with highway congestion by itself.

 

 

I believe you have mass transit confused with personal transit. I am not in favor of major spending on mass transit, it is the most expensive option to build. Automating the roadways would combine our current personal car system with better safety and convenience than mass transit. Automation would obviate the need for the large distances between vehicles to enable an inattentive human driver to "safely" navigate the system. This would allow many more cars to traverse the system over the same period of time. Furthermore, you would never have to stop for traffic signals, slow down for the geezer driving ahead of you, or worry about getting a speeding, drunk driving, or parking ticket. Also, this should save about ten times the number of horrible and preventable deaths (over 40,000) that occurred on 9/11 every single year. Additionally, the system could be powered with renewable sources, like wind and solar, thereby converting about 1/4 of America's total energy use to renewables. All of the technology required to do this is already being used for some application today.

[padren]

I believe you have mass transit confused with personal transit. I am not in favor of major spending on mass transit, it is the most expensive option to build. Automating the roadways would combine our current personal car system with better safety and convenience than mass transit. Automation would obviate the need for the large distances between vehicles to enable an inattentive human driver to "safely" navigate the system. This would allow many more cars to traverse the system over the same period of time. Furthermore, you would never have to stop for traffic signals, slow down for the geezer driving ahead of you, or worry about getting a speeding, drunk driving, or parking ticket. Also, this should save about ten times the number of horrible and preventable deaths (over 40,000) that occurred on 9/11 every single year. Additionally, the system could be powered with renewable sources, like wind and solar, thereby converting about 1/4 of America's total energy use to renewables. All of the technology required to do this is already being used for some application today.

 

I am curious what sort of technologies could automate roadways - it seems easier to automate a flying car system even with current technology (ie we are closer to that) than automate road vehicle highways.

 

With cars densely packed, mechanical failure because a much larger concern, plus factors such as ice, animals, random people - how can an automated system accommodate that in a safe manner? It seems that it would be easier to deal with automating a system in the air, where hazards could be easily identified (birds, wind, ice, other vehicles) against the backdrop of open air, rather than on a road where a puddle hides a pothole the size of a bathtub.

[npts2020]

You accomplish it through imbedded sensors in the road (or above), optical, magnetic, electric or whatever. A central computer tells each vehicle what to do and what route to take once its destination is known. Each vehicle will self monitor and will automatically go for service after going a given number of miles or when detecting below normal performance. Additionally, vehicles should be able to provide feedback to the central computer about road conditions, hopefully being able spot things like potholes before they are large enough to be a problem. For high speeds, the system would have to be enclosed but should be able to go as fast as any bullet train. If you are really concerned about mechanical, guidance, or power failures, it seems to me that any failures would be more catastrophic on a flying system.

[padren]

You accomplish it through imbedded sensors in the road (or above), optical, magnetic, electric or whatever. A central computer tells each vehicle what to do and what route to take once its destination is known. Each vehicle will self monitor and will automatically go for service after going a given number of miles or when detecting below normal performance. Additionally, vehicles should be able to provide feedback to the central computer about road conditions, hopefully being able spot things like potholes before they are large enough to be a problem. For high speeds, the system would have to be enclosed but should be able to go as fast as any bullet train. If you are really concerned about mechanical, guidance, or power failures, it seems to me that any failures would be more catastrophic on a flying system.

 

I find the idea pretty interesting, but I am still pretty skeptical - as far as high speed, you have to plan for very smooth curves and grades, to offset g-force due to acceleration changes even at a constant forward velocity. In order to prevent "adverse surprise conditions" such as falling rocks and such, the infrastructure would have to be largely underground or impact a lot of the nearby terrain.

 

As for failures, if an air vehicle can manage on 3 of 4 engines, if other systems such as com/nav/sensors etc can have redundant components, I don't think it would catastrophic failures would very often at all. I could be way off, but a blown tire seems far more risky in a densely packed automated road system.

 

I am definitely interested in the idea, it just seems really hard to achieve. The degree of road maintenance would increase to quite a costly level I'd imagine.

Are there any good links to conceptual projects of this nature?

[Pangloss]

The funny/sad thing about automated personal vehicles is that their ultimate success or failure would have more to do with human acceptance than actual safety. The system would have an incredibly high expectation of reliability. It can't just be accurate, and in fact it can't even be just better than the current status quo. The fact that it would save 40,000 lives per year would be irrelevant -- the only thing that would matter would be whether or not it killed anybody. For example if it cut the number of annual deaths in half, the story would be that it was killing 20,000 people per year, not that it was saving 20,000. And it's not just a matter of drama-selling-papers, it goes right to the heart of personal freedom. People don't have a sense that they're victims of statistics in auto driving, they feel that their behaviors protect them (which is at least partially true). In an automated system that gets taken away, so the question is raised of whether the system actually makes them safer or really puts them in greater danger.

 

So it would have to be near-perfect -- of a level comparable to US commercial aviation or better (there hasn't been a fatal accident in US comav since Comair 191, 2.5 years ago). And as I understand it the current technology isn't even in that ballpark. Of course we won't really know until such a system is implemented, but from what I've read it needs a lot more attention and funding for a while yet.

Edited December 30, 2008 by Pangloss

[Mokele]

The funny/sad thing about automated personal vehicles is that their ultimate success or failure would have more to do with human acceptance than actual safety.

 

I agree, but would add another layer - there are several possible safety mechanisms which we cannot add to current cars, but which automated cars could have, simply due to human behavior.

 

For instance, imagine some sort of 'super-bumper', capable of safely absorbing damage and restoring itself. If we put those on automated cars, it'll save lives. If we put them on non-automated cars, the death toll will be in the millions because we've just removed the concept of 'consequences' and effectively turned the entire interstate system into the largest game of bumper cars in existence.

 

Basically, by removing human behavior from the equation, we can do a LOT more - have the car computers talk to each other actively within a given distance to ensure safety, knowing traffic light signals far in advance and planning ahead, etc. Plus, no more drunk driving. No more teen drivers. No more old drivers.

 

I agree that human perceptions about loss of control will be an issue, but I also think you underestimate the effectively instantaneous benefits and the possible additional safety measures possible without humans.

 

IMHO, the big problem will be integrating automated vehicles with non-automated ones.

 

Also, IMHO, we should split this thread.

[npts2020]

Well here is the executive summary from a study completed last year for the Texas DoT about this very subject. There are links to the full study (quite a slog to read through but very comprehensive) and entities working on various forms of personal rapid transit. The final conclusion was that all of the technical hurdles would be solved in less than seven years if there is research in that direction.

Pangloss is correct in saying that the main obstacle is acceptance. IMO this is completely irrational and largely due to the fact that few people can visualize how such a system would work. I would think that crashes should be at least as rare as in commercial aviation in a well designed system. The current roadways could be automated as is but other than possibly being safer, would not be much of an improvement. Enclosing the system would enable higher speeds, greater safety, far fewer closures because of weather related problems and less wear and tear from weather. Furthermore, we could upgrade and protect our power grid and other utilities in conjunction with automation and power the whole system with renewable sources.

 

(if you want to attach the last couple of posts to my thread on automated cars, please do so)

[padren]

The funny/sad thing about automated personal vehicles is that their ultimate success or failure would have more to do with human acceptance than actual safety. The system would have an incredibly high expectation of reliability. It can't just be accurate, and in fact it can't even be just better than the current status quo. The fact that it would save 40,000 lives per year would be irrelevant -- the only thing that would matter would be whether or not it killed anybody.

 

That's a good point, and after thinking about it more, I guess human response times to surprise events is really not that great - less than an automated system may be able to achieve.

 

I think though, it may help if we see an increase in "drive by wire" technology, which would have to hit before any automation could be in place anyway. I am sure more than one distracted driver will try to blame their car's UI, and it will take a while to trust it will be more reliable than MS Vista, but after a period of mistrust followed by general acceptance, it could help pave the way for automated systems.

 

Drive by wire is slowly entering the consumer market, I wasn't around for it but I wouldn't doubt it if automatic transmission had it's share of skeptics too. Drive by wire may be the perfect inroad because it doesn't need a high number of adopters to be introduced, whereas automation is only viable if there are a lot of people willing to jump on the bandwagon when it's introduced.

 

Thanks for the link npts2020, I'll be digesting that for a bit.

[npts2020]

Padren; You are welcome. I didn't fully read the executive summary since I read the whole report but I want to be clear that Ms Economides and Mr Longbottom are looking at it from the perspective of having a mixed use (some vehicles not automated) system for a time and not the radical full changeover I am in favor of. IMO the main thing this might achieve is enabling people who are not competent to drive to get on the highways, definitely a good thing but not a much better system than we already have. It may be that is the way it will have to be done but it will be much more expensive in the long run.

Another problem will be the speed with which any system can be built, people will not accept closures on many of the main roads for months and years. Is anyone interested in working on a patent for prefabricated roadways? I see no reason why miles of road cannot be put in place by a single crew every day using this method. It will also facilitate repairs if all you have to do is yank out a section and replace and level it.

[Mokele]

Honestly, mixed-use will be the only option. Even setting aside the technological and logistical difficulties of full conversion, there's the cost - you'd have to basically buy new cars for just about everyone in the US, all at once. Remember, a lot of people cannot afford to just go out and buy a brand new, state-of-the-art car because Uncle Sam says so.

[npts2020]

Honestly, mixed-use will be the only option. Even setting aside the technological and logistical difficulties of full conversion, there's the cost - you'd have to basically buy new cars for just about everyone in the US, all at once. Remember, a lot of people cannot afford to just go out and buy a brand new, state-of-the-art car because Uncle Sam says so.

 

I would think the vast majority of vehicles would be public (I can't imagine myself buying one if there are publicly available ones). The Economides/Longbottom solution to this is the mixed use roadway which is not a bad solution for the short term since older cars could be retrofitted but imo will ultimately make signifcantly improving speed and safety more difficult. At any rate, such a system will take longer to build than your average car's useful life. It would be a big mistake to begin building without having enough cars ready to use the system by the time the first section is completed. In the plan I have envisioned there would be two parts of the system. One part (the high-speed interstate part) would be fully enclosed, the other (the part that goes to everybodies house, workplace, or shopping venue) could be open and possibly have mixed use but would use much lower speeds.

[Sisyphus]

Why would I want an automated, enclosed interstate system like that instead of a mass transit system upgraded to current state of the art and expanded to coverage comparable to the interstate system? That would be enormously expensive, but still, it seems to me, quite a bit less expensive than this proposal. Plus, it wouldn't even require any technology that doesn't exist today, wouldn't provoke public mistrust, would almost certainly be safer, would probably be faster, and would be far more efficient. Am I wrong?

[padren]

I think the "automated enclosed" would allow routing without "transfers" from point to point travel.

 

Though, I personally think a hybrid system would work better because you just can't build a secondary infrastructure large enough to get to everyone's driveway.

[Pangloss]

Why would I want an automated, enclosed interstate system like that instead of a mass transit system upgraded to current state of the art and expanded to coverage comparable to the interstate system?

 

Faster, more personal freedom, etc.

[npts2020]

Well, for one thing mass transit is very inefficient if not heavily utilized. If the vehicles are more the size of a car, it would not make so much difference if you were the only one in the vehicle, plus they wouldn't even run if nobody was using them. Enclosure is going to be necessary for any high-speed transportation. IMO we have more than maxed out our speed limit for safe travel the way our current system is built. I have seen pretty serious and fatal accidents from nothing other than a deer or stray dog wandering onto the interstate. As padren points out you would never have to wait for a transfer, as well. Sisyphus rightly points out the cost for such an undertaking but any option will require trillions of dollars in investment. Upgrading and automating our current roadways is the least expensive option I have seen for any significant improvement of speed, safety, and convenience. What it will one day come down to is are we going to subsidize both a mass transit system and a public road system, cut out one or the other, or convert the public road system to automated personal transit and do away with mass transit altogether? Finally, I would like to know exactly what technology is required that is not available today? AFAIK the only thing required to build a working model is to write the computer algorithms required for control, which requires no new technology, only adapting what we already have.

[Sisyphus]

Alright, I'll play devil's advocate a little longer.

 

Faster, more personal freedom, etc.

 

I doubt faster. Modern high speed trains approach 300mph at peak speeds. So even accounting for waiting for connections (which would almost never be long, if everybody is taking trains - the subway I take to work comes to my stop every three minutes), it would still be much faster than any car (even an automated one) except for very short trips. But if we're talking about replacing the interstate system (and we are, one way or the other), we're not talking about short trips, anyway.

 

I'll grant more personal freedom, if by that you just mean you leave on your own schedule and get taken directly to your destination. But you'd still be forcing people to use this system, and they'd still have to relinquish personal control.

 

Well, for one thing mass transit is very inefficient if not heavily utilized. If the vehicles are more the size of a car, it would not make so much difference if you were the only one in the vehicle, plus they wouldn't even run if nobody was using them.

 

I have to take exception to this. The average train can transport one ton of cargo/passengers 436 miles on one gallon of fuel. That's a few orders of magnitude more efficient than any car. And what train isn't going to be heavily utilized, if they're carrying the bulk of the load currently carried by interstates?

 

Enclosure is going to be necessary for any high-speed transportation.

 

High-speed trains aren't enclosed.

 

As padren points out you would never have to wait for a transfer, as well.

 

True, but as I said above, the more people that take the train, the less the wait. The subways in New York (which is what most people use for travel beyond walking distance) leave every 3 minutes during peak times, and even in the middle of the night come every 10-20 minutes, depending on the line.

 

Which actually brings up another point. Mass transit has a much higher capacity ceiling than cars ever could. In New York the streets and expressways are oversaturated to a degree country folk can't understand without experiencing it, but they're still carrying just a small percentage of total traffic.

 

Sisyphus rightly points out the cost for such an undertaking but any option will require trillions of dollars in investment. Upgrading and automating our current roadways is the least expensive option I have seen for any significant improvement of speed, safety, and convenience.

 

We're both just speculating without enough information, but I don't see how enclosing every interstate could be cheaper than laying track.

 

Finally, I would like to know exactly what technology is required that is not available today? AFAIK the only thing required to build a working model is to write the computer algorithms required for control, which requires no new technology, only adapting what we already have.

 

I don't know what it would require - that's the point. There are no functioning examples, and those would have to be developed and perfected. In fact, there doesn't yet exist even a single self-driving car with adequate capabilities, and it's not for lack of trying. And that's not even counting new types of infrastructure and unforseen problems.

[npts2020]

Alright, I'll play devil's advocate a little longer.

 

 

 

I doubt faster. Modern high speed trains approach 300mph at peak speeds. So even accounting for waiting for connections (which would almost never be long, if everybody is taking trains - the subway I take to work comes to my stop every three minutes), it would still be much faster than any car (even an automated one) except for very short trips. But if we're talking about replacing the interstate system (and we are, one way or the other), we're not talking about short trips, anyway.

 

I'll grant more personal freedom, if by that you just mean you leave on your own schedule and get taken directly to your destination. But you'd still be forcing people to use this system, and they'd still have to relinquish personal control.

 

 

 

I have to take exception to this. The average train can transport one ton of cargo/passengers 436 miles on one gallon of fuel. That's a few orders of magnitude more efficient than any car. And what train isn't going to be heavily utilized, if they're carrying the bulk of the load currently carried by interstates?

 

 

 

High-speed trains aren't enclosed.

 

 

 

True, but as I said above, the more people that take the train, the less the wait. The subways in New York (which is what most people use for travel beyond walking distance) leave every 3 minutes during peak times, and even in the middle of the night come every 10-20 minutes, depending on the line.

 

Which actually brings up another point. Mass transit has a much higher capacity ceiling than cars ever could. In New York the streets and expressways are oversaturated to a degree country folk can't understand without experiencing it, but they're still carrying just a small percentage of total traffic.

 

 

 

We're both just speculating without enough information, but I don't see how enclosing every interstate could be cheaper than laying track.

 

 

 

I don't know what it would require - that's the point. There are no functioning examples, and those would have to be developed and perfected. In fact, there doesn't yet exist even a single self-driving car with adequate capabilities, and it's not for lack of trying. And that's not even counting new types of infrastructure and unforseen problems.

 

Feel free to play devils advocate I am interested in any arguments someone could put forth.

 

First of all, there is no rail that goes even close to 300mph (record American trial speed is 295 kph-see wikipedia/high speed rail) in the United States. Furthermore, it is doubtful that one could run trains that fast on the current infrastructure (crossing and signal issues, current condition of rails, etc.). Every high speed rail I have seen is separated from everything else by being elevated, having a fence, or being underground in all places they operate at near max speeds. I would like to see your source for the mpg of a rail because I can't believe a train pulling as few as only three or four cars could even get up to speed using any acceleration on a gallon of gas much less go 400+ miles after. I understand that rush hour mass transit in highly urbanized areas is by far the most efficient transportation option, and should be continued in many places. Unfortunately that is only a fraction of the miles travelled on an annual basis by Americans. People who live near public transit and only basically only travel to and from work generally do not own cars, that is not the vast majority of us. Also, on an automated system you would not have to get youself to wherever public transit is then get to where you want to go from wherever it ends. An automated car would show up in your driveway and take you to the driveway of wherever you were going. While it is probable that some will not like "giving up control", that sentiment is totally illogical. We "give up control" every time we take a cab, fly on a commercial plane, take public transportation, or ride with a buddy to go play golf. So far as "forcing" anyone to take the system, I really don't see your point since I am not advocating making cars illegal. Right now I am "forced" to use a car because there is no viable alternative where I live, how is having a choice of one or the other forcing anyone to do anything? In order to not write a book here, I will finish by saying that cost of various methods of automation is a subject I have been pursuing for over six months now. As you say there is no working models at present (of the sort I advocate leastways), this makes getting good cost analysis information very difficult. Suffice it to say, it is not complete speculation when I claim that the system I have envisioned will be less expensive, in the long run, than high speed rail and if it is powered with wind and solar energy, efficiency becomes less of an issue.

[Mr Skeptic]

I have to take exception to this. The average train can transport one ton of cargo/passengers 436 miles on one gallon of fuel. That's a few orders of magnitude more efficient than any car. And what train isn't going to be heavily utilized, if they're carrying the bulk of the load currently carried by interstates?

 

Wow! I couldn't believe that until I looked it up, but it seems to be true. I'm guessing that is a result of the train's shape, as well as not having to stop. But I don't know if that would apply to mass-transit trains, since these will be lighter, shorter, and stop more. I need to look that up.

[npts2020]

Also, an automated system could use rails, but that will add significant cost to the system and likely rule out any possibility of mixed use until conversion is completed (not necessarily a bad thing IMO).

[Baby Astronaut]

...I can't believe a train pulling as few as only three or four cars could even get up to speed using any acceleration on a gallon of gas much less go 400+ miles after.

Check out an investigation by FactCheck.org

 

Each year the railroads are required to submit reports to the federal Surface Transportation Board, the regulatory body that took over some of the functions of the old Interstate Commerce Commission. The annual reports of each railroad are public information, available on the STB's Web site. Buried amid all the facts about the number of railroad ties replaced, cubic yards of ballast placed and the cost of new locomotives, the railroads also report totals for the number of gallons of diesel fuel consumed and tons of freight moved. The government doesn't tally up those figures anymore, but the Association of American Railroads does. And now, we have done the same.

[npts2020]

Thanks for the link BabyAstronaut. I see that passenger trains were excluded and they mention a few other negative things about using locomotives but it does seem they are the most efficient means of going from point A to point B that we now have available. My problem with the efficiency of railroads has never been on the fuel side (other than running them empty) so much as limited range (time efficiency in getting to where you want to go). You are not likely to ever have a train stop in front of your house or place of business if there is not one there already, whereas personal transit could be made to go anywhere current roads go without major changes to infrastructure. To finish I would just like to say that, railroads mostly still burn fossil fuels with no plan to change. The system I advocate would use solar and wind power, making efficiency less of a concern.

[CaptainPanic]

Passenger trains should be excluded, because they're less efficient. (They carry chairs and other things for comfort (and even a toilet!), they have lights and airco, etc... not likely to be equally efficient).

 

And you should compare it to a truck, not a normal car.

 

But I do believe that ships are even more efficient for transporting cargo, even upstream on major rivers. (Sorry, no link to back up my claim).

[npts2020]

CaptainPanic; I have always heard water transport was the most efficient, also. Problem is that is even more limited than railways for where it can usefully go. Also passenger trains are excluded from those numbers.

[CaptainPanic]

Hehe, ships are more limited in your country perhaps. There are some places where it does work. Click for an aerial picture of Amsterdam. But I agree with npts2020. Ships should only be used where possible, and the options are limited.

 

Regarding the personal automated transit, I have two points to make...

 

First, in case of a completely personalized system, you'll be needing the same grid as our normal roads... which simply means you'll have lots of corners and crossings like we have today.

 

If the grid is the same, then the paths of (automated) vehicles will certainly cross. This brings a danger, which is now covered by traffic lights and traffic laws. If cars won't stop for a crossing, then you either have to build a bridge (highway style) or weave the twelvestreams of cars such that they never collide. Sorry, but I really doubt any engineer would ever attempt to design this. Not with real living people in the vehicles. For cargo, perhaps it can be done... (Twelve streams of cars: two roads, traffic in two directions is 4 streams arriving at the crossroad. Then each stream can go in 3 directions: left, right and straight).

 

A chance that there is a break down will be almost unacceptable at increased speeds and densely packed roads... Therefore, I suggest that much more bridges, flyovers, or whatever elevated roads need to be built...

 

The point I'm making here is that it's not going to fit in our cities... you cannot automate each and every crossroad using bridges and tunnels. You would not have to rebuild the roads, but the entire country.

 

Second, compared to a train, a personal vehicle will have more air resistance. This will increase fuel consumption... Unless people can have some G-forces, you need to slow down and speed up again at every corner... same as now. There is no increased efficiency in this idea.

 

I do think that automating cars can have benefits... but I would study this only for the highways (or interstate roads or whatever you call them - those roads for longer distances where you can drive fast).