Go Back  Bike Forums > Bike Forums > Advocacy & Safety
Reload this Page >

Would a self driving car world make it safe for cyclists?

Notices
Advocacy & Safety Cyclists should expect and demand safe accommodation on every public road, just as do all other users. Discuss your bicycle advocacy and safety concerns here.

Would a self driving car world make it safe for cyclists?

Old 11-28-17, 08:44 PM
  #701  
Arizona Dessert
 
noisebeam's Avatar
 
Join Date: Jun 2004
Location: AZ
Posts: 15,029

Bikes: Cannondale SuperSix, Lemond Poprad. Retired: Jamis Sputnik, Centurion LeMans Fixed, Diamond Back ascent ex

Liked 2,169 Times in 1,288 Posts
Originally Posted by Ninety5rpm
But the main thing is that the luxury of being able to work or watch a move or surf the web or nap instead of driving will outweigh the "joy" of driving, especially in commuter traffic, for the vast majority of commuters.
"Would you like me to drive around the block so you can finish episode 3?"

"Yeah. Can you always do that and not interrupt me next time?"
noisebeam is offline  
Old 11-28-17, 08:46 PM
  #702  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by tyrion
I disagree with this prediction because it would be a backward step environmentally to push pedestrians and bicyclists off the road (not to mention extremely disruptive) so that political argument would be difficult to make.

There's a big windfall of benefits that will happen when every vehicle is smart, communicating and cooperating, but to get there they'd have to push every existing vehicle off the road and I don't see that happening for decades.

I predict there will be new smart-vehicle-only roads that will enable the windfall of benefits, and the new smart roads will be alongside (or more likely above) the existing anything goes roads. These new roads will be much higher capacity because higher speeds will be viable, and stops and congestion will be minimized (maybe eliminated) by smart cooperation, so one lane could handle the traffic of 4 or more conventional lanes. The new smart roads will be a faster way to travel than existing "open", archaic roads and this will encourage people to use smart cars, which in turn will compel the development of more smart roads.
I think the advantages of traveling on a smart-only road over an open road will be marginal at best. While I do believe AVs will be much much less prone to error than humans, we're already well into the diminishing returns end of the vehicular traffic efficiency spectrum with human drivers. Again, smart cars can make marginal movements, and some more marginal improvements if it's only smart cars on the "smart" roads, but if you set up a computer simulation of traffic in a congested city, you're not going to get much different overall throughput if you program it with computer drivers rather than human drivers.

I know the biggest difference is supposed to be at intersections where if you have two smart streets intersecting with only smart cars, and human-driven cars as well as pedestrians and bicyclists banned, then in theory you should be able to get them through the intersection more efficiently with car-to-car communications and tightly controlled timing than with traffic signals.

But multilane roundabouts and traffic circles should be able to accomplish almost the same results (in terms of throughput), and they don't require the banning of humans, so I think that's much more likely. Besides, once you reach a certain level of congestion, traffic signals are simply the most efficient way to go anyway. Then it doesn't matter at all if the drivers are human or computer.

Last edited by Ninety5rpm; 11-28-17 at 08:55 PM.
Ninety5rpm is offline  
Old 11-28-17, 08:49 PM
  #703  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by noisebeam
"Would you like me to drive around the block so you can finish episode 3?"

"Yeah. Can you always do that and not interrupt me next time?"


Ninety5rpm is offline  
Old 11-28-17, 08:54 PM
  #704  
Senior Member
 
Join Date: Jul 2012
Location: Canada, PG BC
Posts: 3,849

Bikes: 27 speed ORYX with over 39,000Kms on it and another 14,000KMs with a BionX E-Assist on it

Liked 57 Times in 49 Posts
Originally Posted by Ninety5rpm
I think the advantages of traveling on a smart-only road over an open road will be marginal at best. While I do believe AVs will be much much less prone to error than humans, we're already well into the diminishing returns end of the vehicular traffic efficiency spectrum with human drivers. Again, smart cars can make marginal movements, and some more marginal improvements if it's only smart cars on the "smart" roads, but if you set up a computer simulation of traffic in a congested city, you're not going to get much different overall throughput if you program it with computer drivers rather than human drivers.
I would beg to differ, if all those computer driven vehicles are operated by a connected computer "network"... THEN the flow of traffic can/could/would , flow more fluently, thus the congestion and parking "problem" would ease, if most personal vehicles would be taken off the road...
350htrr is offline  
Old 11-28-17, 08:59 PM
  #705  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by 350htrr
I would beg to differ, if all those computer driven vehicles are operated by a connected computer "network"... THEN the flow of traffic can/could/would , flow more fluently, thus the congestion and parking "problem" would ease, if most personal vehicles would be taken off the road...
If you look at an aerial view video of congested Los Angeles I don't see what difference it would make to simply replace all those human-driven cars with computer-driven cars. This is congestion on freeways where all the intersection conflicts are already mitigated with separated grades, everywhere.

And if you're thinking of Manhattan, well now we're back to banning pedestrians, bicyclists and human-driven cars. That's not going to happen, even if it would help.

Last edited by Ninety5rpm; 11-28-17 at 09:03 PM.
Ninety5rpm is offline  
Old 11-28-17, 09:20 PM
  #706  
C*pt*i* Obvious
 
SHBR's Avatar
 
Join Date: Dec 2013
Location: Shanghai
Posts: 1,337
Liked 53 Times in 44 Posts
Food for thought.

SHBR is offline  
Old 11-28-17, 09:28 PM
  #707  
Senior Member
 
tyrion's Avatar
 
Join Date: Oct 2015
Location: San Diego, California
Posts: 4,076

Bikes: Velo Orange Piolet

Liked 2,012 Times in 972 Posts
Originally Posted by Ninety5rpm
I think the advantages of traveling on a smart-only road over an open road will be marginal at best.
Smart-only roads can be much faster when all the traffic is communicating and cooperating and there's an assurance that there are no "dumb" things (pedestrians, bikes, dogs, deer, etc.) on the road. If there are sensors (in the road itself or on tiny drones above) that can signal "all clear ahead" then the speed limit can be very high. And cars will be able to form into tight convoys for aero efficiency.
tyrion is offline  
Old 11-28-17, 09:50 PM
  #708  
Senior Member
 
tyrion's Avatar
 
Join Date: Oct 2015
Location: San Diego, California
Posts: 4,076

Bikes: Velo Orange Piolet

Liked 2,012 Times in 972 Posts
Originally Posted by Ninety5rpm
If you look at an aerial view video of congested Los Angeles I don't see what difference it would make to simply replace all those human-driven cars with computer-driven cars.
It'd make a massive difference. Most of the slowdowns are caused by cars being forced to merge, which computers can do a billion times better than people can. Merging 4 lanes into 3 would cripple a human driven system but it would only slow down a computer driven system by 25% at most. It's the density of the traffic that causes the slowdown - a single car doesn't even have to slow down when 4 lanes merge to 3 - so faster traffic mean less density, which is in turn less affected by merges. Not only that, the speed transitions would be smooth and not herky jerky like human traffic.
tyrion is offline  
Old 11-28-17, 10:00 PM
  #709  
Senior Member
 
tyrion's Avatar
 
Join Date: Oct 2015
Location: San Diego, California
Posts: 4,076

Bikes: Velo Orange Piolet

Liked 2,012 Times in 972 Posts
Originally Posted by Ninety5rpm
And if you're thinking of Manhattan, well now we're back to banning pedestrians, bicyclists and human-driven cars. That's not going to happen, even if it would help.
Manhattan might be screwed. Until we get hovercars. Or jetpacks.
tyrion is offline  
Old 11-28-17, 10:47 PM
  #710  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by tyrion
Manhattan might be screwed. Until we get hovercars. Or jetpacks.
I think Manhattan will benefit greatly from AVs, and the efficient carpooling they enable.
Ninety5rpm is offline  
Old 11-28-17, 10:56 PM
  #711  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by tyrion
It'd make a massive difference. Most of the slowdowns are caused by cars being forced to merge, which computers can do a billion times better than people can. Merging 4 lanes into 3 would cripple a human driven system but it would only slow down a computer driven system by 25% at most. It's the density of the traffic that causes the slowdown - a single car doesn't even have to slow down when 4 lanes merge to 3 - so faster traffic mean less density, which is in turn less affected by merges. Not only that, the speed transitions would be smooth and not herky jerky like human traffic.
It's not that simple. Again, in a computer simulation of traffic and merging, even if you use theoretically optimal drivers you get gridlock. Throughput is throughput. And human driver merging behavior is not that far off from theoretically optimal ones. Sure zippers could work better, but these are marginal improvements - human drivers get pretty far down the road of diminishing returns. You're not going to get an order of magnitude improvement - not even close. Not even double. 10% more efficient? Maybe.
Ninety5rpm is offline  
Old 11-28-17, 11:33 PM
  #712  
Senior Member
 
tyrion's Avatar
 
Join Date: Oct 2015
Location: San Diego, California
Posts: 4,076

Bikes: Velo Orange Piolet

Liked 2,012 Times in 972 Posts
Originally Posted by Ninety5rpm
Again, in a computer simulation of traffic and merging, even if you use theoretically optimal drivers you get gridlock.
No, gridlock (to the point of stopping) is only the result of gross human inefficiency (or complete road blockage).

And human driver merging behavior is not that far off from theoretically optimal ones.
Not even close. All the speeding up and slowing down and the stretching and condensing of gaps between cars is slop in the system that can mostly be optimized away with computing power. Losing 1 out of 4 lanes will require a 25% slowdown in an optimized system (at full capacity - in sparse traffic less slowdown is required). I don't know how close a smart system could get to the ideal in the real world, but a human system can't get anywhere near that ideal. It's not that humans are that dumb, it's that communications between drivers is so limited.
tyrion is offline  
Old 11-28-17, 11:51 PM
  #713  
Senior Member
 
tyrion's Avatar
 
Join Date: Oct 2015
Location: San Diego, California
Posts: 4,076

Bikes: Velo Orange Piolet

Liked 2,012 Times in 972 Posts
Originally Posted by Ninety5rpm
You're not going to get an order of magnitude improvement - not even close. Not even double.
I guarantee a ten fold increase within 40 years. That is one lane of smart road will have 10 times the capacity of one human driven lane. That's not just because the smart road traffic will be faster overall, but because it can also be much denser.

The technological part isn't that difficult, but it's going to take some time for the right standards to shake out and solidify, and it will take time to build the infrastructure.
tyrion is offline  
Old 11-29-17, 12:09 AM
  #714  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by tyrion
No, gridlock (to the point of stopping) is only the result of gross human inefficiency (or complete road blockage).



Not even close. All the speeding up and slowing down and the stretching and condensing of gaps between cars is slop in the system that can mostly be optimized away with computing power. Losing 1 out of 4 lanes will require a 25% slowdown in an optimized system (at full capacity - in sparse traffic less slowdown is required). I don't know how close a smart system could get to the ideal in the real world, but a human system can't get anywhere near that ideal. It's not that humans are that dumb, it's that communications between drivers is so limited.
If it's not even close, then why is computer network traffic flow so similar to human-driven vehicular traffic flow?

What is the basis for the claim that the slop due to inefficient speeding up and slowing down is as significant to overall flow as you seem to believe?
Ninety5rpm is offline  
Old 11-29-17, 07:55 AM
  #715  
genec
 
genec's Avatar
 
Join Date: Sep 2004
Location: West Coast
Posts: 27,079

Bikes: custom built, sannino, beachbike, giant trance x2

Liked 4,532 Times in 3,158 Posts
Originally Posted by tyrion
I disagree with this prediction because it would be a backward step environmentally to push pedestrians and bicyclists off the road (not to mention extremely disruptive) so that political argument would be difficult to make.

There's a big windfall of benefits that will happen when every vehicle is smart, communicating and cooperating, but to get there they'd have to push every existing vehicle off the road and I don't see that happening for decades.

I predict there will be new smart-vehicle-only roads that will enable the windfall of benefits, and the new smart roads will be alongside (or more likely above) the existing anything goes roads. These new roads will be much higher capacity because higher speeds will be viable, and stops and congestion will be minimized (maybe eliminated) by smart cooperation, so one lane could handle the traffic of 4 or more conventional lanes. The new smart roads will be a faster way to travel than existing "open", archaic roads and this will encourage people to use smart cars, which in turn will compel the development of more smart roads.
I doubt we will see a massive build up of mew roads for AVs... as it is we do not economically do well with our existing infrastructure.

But I do suspect you may be somewhat correct, in that existing carpool lanes may be "redesigned... " if AVs prove to be safe and efficient enough.
genec is offline  
Old 11-29-17, 09:25 AM
  #716  
Been Around Awhile
 
I-Like-To-Bike's Avatar
 
Join Date: Oct 2004
Location: Burlington Iowa
Posts: 30,056

Bikes: Vaterland and Ragazzi

Liked 1,606 Times in 1,084 Posts
Originally Posted by genec
I doubt we will see a massive build up of mew roads for AVs... as it is we do not economically do well with our existing infrastructure.

But I do suspect you may be somewhat correct, in that existing carpool lanes may be "redesigned... " if AVs prove to be safe and efficient enough.
How mean you are, injecting a dose of economic reality into this heap of steaming technological daydreaming!

And to top it off you applied a forbidden word "If" into a discussion where any and all tests and simulations prove that the theory is correct, and in fact that just conducting such tests and simulations prove that the sought after objective of a safe and efficient product is already at hand.
I-Like-To-Bike is offline  
Old 11-29-17, 10:15 AM
  #717  
Senior Member
 
Join Date: Sep 2013
Location: Massachusetts
Posts: 4,530
Liked 663 Times in 443 Posts
Originally Posted by noisebeam
Livestock, horses, equestrians on the highway. I encounter all on my cycle commute.


You also CLAIM to encounter autonomous vehicles, so what are we to believe?

Think outside the bun? Ride inside the Bell

-mr. bill
mr_bill is offline  
Old 11-29-17, 10:32 AM
  #718  
genec
 
genec's Avatar
 
Join Date: Sep 2004
Location: West Coast
Posts: 27,079

Bikes: custom built, sannino, beachbike, giant trance x2

Liked 4,532 Times in 3,158 Posts
Originally Posted by I-Like-To-Bike
How mean you are, injecting a dose of economic reality into this heap of steaming technological daydreaming!

And to top it off you applied a forbidden word "If" into a discussion where any and all tests and simulations prove that the theory is correct, and in fact that just conducting such tests and simulations prove that the sought after objective of a safe and efficient product is already at hand.
Try not to get too excited... we are talking about different economic sectors here... the technology is being pushed by private companies, using private investor funds.

Roads, on the other hand, come from public monies; taxes and grants... none of which is even maintaining our current infrastructure.
genec is offline  
Old 11-29-17, 10:43 AM
  #719  
Arizona Dessert
 
noisebeam's Avatar
 
Join Date: Jun 2004
Location: AZ
Posts: 15,029

Bikes: Cannondale SuperSix, Lemond Poprad. Retired: Jamis Sputnik, Centurion LeMans Fixed, Diamond Back ascent ex

Liked 2,169 Times in 1,288 Posts
Originally Posted by mr_bill
You also CLAIM to encounter autonomous vehicles, so what are we to believe?
I encounter far more autonomous vehicles. Usually 2-3 each morning and 2-3 on the way home. Never less than 2 a day. A couple of Saturday's ago I saw 14 during one 3hr ride. I suspect some were the same vehicle.

I encounter equestrian about 2x/yr. They ride in the bike lane usually - getting from their county horse property to the trails in the park and back. I had to follow some cattle being driving along the road about 3yr ago. Every once in a while a stray cow is found on an urban freeway.
Capturecow.JPG
noisebeam is offline  
Old 11-29-17, 10:45 AM
  #720  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by genec
Try not to get too excited... we are talking about different economic sectors here... the technology is being pushed by private companies, using private investor funds.

Roads, on the other hand, come from public monies; taxes and grants... none of which is even maintaining our current infrastructure.
Yeah, AV technology is developing and arriving at 100 - 1000 times the speed infrastructure is improved/expanded.

Counting on infra improvements to reap the benefits of AV technology is probably even more silly than counting on infra improvements to make cycling safer.
Ninety5rpm is offline  
Old 11-29-17, 10:46 AM
  #721  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by noisebeam
I encounter far more autonomous vehicles. Usually 2-3 each morning and 2-3 on the way home. Never less than 2 a day. A couple of Saturday's ago I saw 14 during one 3hr ride. I suspect some were the same vehicle.
Attachment 590664
How often do you get a good enough look to ascertain whether the human driver is engaged or passive?
Ninety5rpm is offline  
Old 11-29-17, 11:08 AM
  #722  
Senior Member
 
Join Date: Jul 2012
Location: Canada, PG BC
Posts: 3,849

Bikes: 27 speed ORYX with over 39,000Kms on it and another 14,000KMs with a BionX E-Assist on it

Liked 57 Times in 49 Posts
Originally Posted by Ninety5rpm
If you look at an aerial view video of congested Los Angeles I don't see what difference it would make to simply replace all those human-driven cars with computer-driven cars. This is congestion on freeways where all the intersection conflicts are already mitigated with separated grades, everywhere.

And if you're thinking of Manhattan, well now we're back to banning pedestrians, bicyclists and human-driven cars. That's not going to happen, even if it would help.
I am thinking of the difference just at the lights... The light turns green, the first person starts off, when the first car starts to move the second car starts to move and so on, a delay of about a second per human driven car... With autonomous vehicles they ALL would start moving at the same instant keeping the same distance between themselves.. IMO that in itself would double any roads capacity...
350htrr is offline  
Old 11-29-17, 11:12 AM
  #723  
Banned
 
Join Date: Apr 2008
Posts: 1,341
Likes: 0
Liked 0 Times in 0 Posts
Originally Posted by 350htrr
I am thinking of the difference just at the lights... The light turns green, the first person starts off, when the first car starts to move the second car starts to move and so on, a delay of about a second per human driven car... With autonomous vehicles they ALL would start moving at the same instant keeping the same distance between themselves.. IMO that in itself would double any roads capacity...
Ah. Interesting. What I see at congested traffic signal stops is perhaps some inefficiency with the first few cars when the light first turns green, but within a few cars the traffic is flowing into the intersection at near capacity rates, so i don't think there is much room for improvement there. I'd like to see it modeled.
Ninety5rpm is offline  
Old 11-29-17, 11:39 AM
  #724  
Senior Member
 
Join Date: Sep 2013
Location: Massachusetts
Posts: 4,530
Liked 663 Times in 443 Posts
Originally Posted by Ninety5rpm
How often do you get a good enough look to ascertain whether the human driver is engaged or passive?
It's pretty easy to see when walking (and I'd imagine when biking).

When the Uber folks are driving, they have their hands at 10-2. When monitoring, they have their open hands palms up just below the steering wheel at about 8-4. Looks kind of like they are meditating with their eyes open to be honest.

-mr. bill
mr_bill is offline  
Old 11-29-17, 11:47 AM
  #725  
Arizona Dessert
 
noisebeam's Avatar
 
Join Date: Jun 2004
Location: AZ
Posts: 15,029

Bikes: Cannondale SuperSix, Lemond Poprad. Retired: Jamis Sputnik, Centurion LeMans Fixed, Diamond Back ascent ex

Liked 2,169 Times in 1,288 Posts
Originally Posted by Ninety5rpm
How often do you get a good enough look to ascertain whether the human driver is engaged or passive?
Often. They are usually passive. No hands on wheel. Watching me in rear view while moving forward or turning, etc.
I am on the lookout for one without anyone in driver seat. Haven't seen yet.
noisebeam is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Your Privacy Choices -

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.