How Much Road Capacity is Wasted Due to Poor Driving?

>
> Uh, that won't be driving anymore. That would be riding.
> --------------
> Alex
>

I'd actually prefer it that way, most of the time. I know an extra 30
minutes of sleep each morning would be nice. Next logical step . . .
sleeper cars. Go to sleep in your work clothes and literally wake up when
you reach the office. -Dave

Scott en Aztl=E1n wrote:
> Anyone want to take a guess as to how much of our existing road
> capacity is wasted by LLBs, poor mergers, Sloths, rubberneckers,
> people who don't speed up when the traffic ahead of them does, etc.
> etc.?
>
> I'd say the waste is at least 40%, i.e. that our roads never achieve
> more than 60% of their true capacity due to incompetent drivers.

People should be taught to drive more aggressively.

One time, I was driving in fairly heavy but still free-flowing
traffic (middle lane of 3 going at 80km/h, fast lane at 90km/h).
I was in the middle lane, and we passed an on-ramp (traffic
merges into the slow lane) where the merged traffic was doing
about 50km/h. A car that had just merged had its indicator on.

So the woman driving in front of me in my lane (about 10m behind
the driver in the adjacent lane who was indicating) slammed on
her brakes, going down from 80 to 35 nearly instantly.
This, apart from nearly causing me to crash, caused a big
middle-lane clump. It even caused people in the fast lane to
slow down (let's not even begin to try and understand the
mentality of that).

Fortunately, traffic was not quite so heavy as usual, and it
managed to recover. But it occurred to me that this sort of
event is exactly how traffic congestion starts every weekday...
when traffic is heavy enough that the clump and the 35km/h
becomes the upper limit until there is a let-up several hours later.

The worst thing is that the woman probably had no idea just
how dangerous her manoeuvre was, and had no idea that she
could have just caused 50,000 people to waste an extra hour
of their time than if she hadn't done that.

If questioned on the matter, she would have said that she
was being considerate and letting the other driver in.
If it were pointed out that she were being considerate to
1 person and inconsiderate to several people behind her,
she would dismiss the people behind her as being selfish.

It isn't rocket science to know that traffic flows best when
the people who have the right of way, actually use it. If some
jerk tries to move into my lane when I have right of way,
I don't move out of the way. If the guy wants to write off his
car and pay for me to get a new car, he can be my guest.
Funnily enough, nobody has taken me up on that offer yet.

Dave C. wrote:

> But a self-driving car could be really EASY to make, in comparison, with
> less electronic gadgetry than that which powers most modern calculators.
> The real hurdle is building the infrastructure needed to support
> self-driving vehicles. Like retro-fitting all roadside signs (even the mile
> markers) with RFID or something that the car can sense. Then installing
> sensors in the roads and bridges to allow automated steering (something to
> aim at, or steer AWAY from). ETC. It will be a major headache. But the
> car itself, will be really rather simple, as long as it doesn't have to
> conform to DARPA.

How about every road be fitted with a set of guide rails, and cars
have wheels modified to fit onto those guide rails?

With a central computer controller, there would be no need for more
than one lane of active traffic per direction (and perhaps a backup
lane in case of a blockage). The only reason we need multi-lane roads
is because people do not drive at the optimal speed, for fear of the
person in front of them doing something unexpected (or because they
are lazy).

This would alleviate any problems such as lane changes.
Similarly, intersections would be smoother: fewer in number
due to more efficient route selecting by the computer, and
no need for blanket red lights, since trains^H^H^Hffic would all
be controlled by the computer.

> This would alleviate any problems such as lane changes.
> Similarly, intersections would be smoother: fewer in number
> due to more efficient route selecting by the computer, and
> no need for blanket red lights, since trains^H^H^Hffic would all
> be controlled by the computer.


You're on the right "track", so to speak.

A computer controlled system would create a virtual train, but without
rails. It would be more sophisticated in that it would allow for
asynchronous additions / deletions of cars at arbitrary intervals, and it
would be the ultimate in efficient merging.

In article .com>, Old Wolf wrote:

> So the woman driving in front of me in my lane (about 10m behind
> the driver in the adjacent lane who was indicating) slammed on
> her brakes, going down from 80 to 35 nearly instantly.
> This, apart from nearly causing me to crash, caused a big
> middle-lane clump. It even caused people in the fast lane to
> slow down (let's not even begin to try and understand the
> mentality of that).

I can explain why. When someone does something stupid like that in the
middle lane, there are two choices, speed up or slow down. If I'm rather
close to the person it's usually best to speed up and get away before
they do the next dumb thing. However, if behind and further back, slowing
is the best option because one has to be ready for the next stupid move
or someone avoiding the person who made the stupid move. Not to mention
people who take risks to get by someone who is needlessly slowing traffic.
(aka scott's sloth kills)


> Fortunately, traffic was not quite so heavy as usual, and it
> managed to recover. But it occurred to me that this sort of
> event is exactly how traffic congestion starts every weekday...
> when traffic is heavy enough that the clump and the 35km/h
> becomes the upper limit until there is a let-up several hours later.

Exactly. It's these things that cause the jams and reduce travel speeds.

> If questioned on the matter, she would have said that she
> was being considerate and letting the other driver in.
> If it were pointed out that she were being considerate to
> 1 person and inconsiderate to several people behind her,
> she would dismiss the people behind her as being selfish.

Exactly. This is a form of what I call "consideration" in quotes. It puts
the needs of one person over those of everybody else. Usually it requires
one selfish person trying to jam their way in and one enabler who lets
them. However, in some instances, just the enabler. I see the later
usually when I am using the bicycle. Someone will stop and wave me
through. I didn't want them to do, they just did it, screwing up the
flow, my timing, etc because they had to be "considerate".

Jim Yanik .> wrote:

: > who squeeze people out who are trying to merge,
:
: These are part of the PROBLEM;they screw up and expect others to
: slow,or stop,and make way for them,CAUSING slower traffic flow.

I have seen unproductive behavior from both, those who are trying to
merge and those already in the "main" lane.

Some people merge too soon, thus wasting a lot of road space in front
and slowing down the process. Others try to gain a few car lengths by
continuing without merging even beyond the obvious merger point.

Some drivers in the main lane leave that lane and try to merge back a
few cars ahead. Others won't let a merging car in. Etc.

Old Wolf > wrote:

: One time, I was driving in fairly heavy but still free-flowing
: traffic (middle lane of 3 going at 80km/h, fast lane at 90km/h).
: I was in the middle lane, and we passed an on-ramp (traffic
: merges into the slow lane) where the merged traffic was doing
: about 50km/h. A car that had just merged had its indicator on.
:
: So the woman driving in front of me in my lane (about 10m behind
: the driver in the adjacent lane who was indicating) slammed on
: her brakes, going down from 80 to 35 nearly instantly.
: This, apart from nearly causing me to crash, caused a big
: middle-lane clump. It even caused people in the fast lane to
: slow down (let's not even begin to try and understand the
: mentality of that).
:
: Fortunately, traffic was not quite so heavy as usual, and it
: managed to recover. But it occurred to me that this sort of
: event is exactly how traffic congestion starts every weekday...
: when traffic is heavy enough that the clump and the 35km/h
: becomes the upper limit until there is a let-up several hours later.
:
: The worst thing is that the woman probably had no idea just
: how dangerous her manoeuvre was...

Not saying your feelings are wrong in this particular case but also
think of the other side in similar sits: I can be aggressive but I
don't know what somebody else is going to do. Some people merge
without looking, at slow speed, and take their sweet time accelerating.
People behind me may be annoyed, but I am the one who would have
colloided with or rear-ended the merging car.

They may not see everything I see. It really is their job to keep
adequate distance and be alert enough to react to my brakes.

For that matter, I have had people honking at me when I am
stopped at a stop sign and there is a small child crossing in front
of my car! The don't see the child and somehow think of a stop
sign as a place where you gently touch the brake pedal once and
move on.

Anyway, if you are behind me, please remember that you don'tsee
what I see and you don't face the same risk that I do.

In article ws.net>,
Dave C. > wrote:
>>
>> A human driver performs the same task, without all that much trouble.
>> The DARPA rules aren't unreasonable.
>
>
>A human mind is a very efficient super computer. It can process a LOT of
>information very quickly. Not all humans use their brains efficiently, but
>it doesn't matter. A human brain can still process a lot of information
>quickly.

Yep.

>You are right that the DARPA rules aren't unreasonable . . . for the purpose
>of the DARPA "race". The DARPA challenge is to create an autonomous weapon
>of some sort capable of self-locomotion and self-navigation. Because it is
>a WEAPON, it must operate alone without help from anything outside (because
>it is in enemy territory obviously, so it would have to MISTRUST anything
>outside trying to help to guide it, if such existed)

And drivers operate in a basically similar environment. The big
difference is we generally stick to the roads, but that's not
sufficient to make the problem tractable.

>Now compare that with the challenge of building a self-driving car. The
>self-driving car doesn't have to operate with no contact with the world
>around it. While you theoretically could take any car into a hostile
>environment, your average commuter vehicle would encounter nothing more
>hostile than automated toll collection booths.

Except, perhaps, weather. Worn and damaged road surfaces. Objects and
animals on the road. And the big one, not in the DARPA test: OTHER
VEHICLES.

>DARPA is looking for a very advanced form of self-driving vehicle, when a
>plain old ordinary commuter car that can drive itself hasn't even been
>invented yet. (DARPA is skipping a few intermediate, easier steps.) -Dave

On the contrary; DARPA is making things much easier by not trying to
make the self-driving vehicle capable of dealing with traffic,
something a commuter car would have to do.
--
There's no such thing as a free lunch, but certain accounting practices can
result in a fully-depreciated one.

"Dave C." wrote:
>
> > Let me get this straight, you think that on average our roads are only
> > carrying 5% of their theoretical maximum capacity because of incompetent
> > drivers?
>
> Yes, because incompetent drivers create a snowball effect. Good examples
> (just a couple out of millions) . . . the sloth ahead of you is doing 10
> below the limit in good weather in an area with NO PASSING ZONES. So that
> slows you down just a little, right? Well, slowing down just a little
> causes you to get stuck behind a *$#)*$)#* SCHOOL BUS FOR THE NEXT FIFTEEN
> MILES MAKING STOPS EVERY FEW FEET!!!! Same with traffic. Many people with
> commutes longer than five minutes have noticed that if they leave even five
> minutes later, it can DOUBLE their commute time (or worse). Or, if they get
> stuck behind an LLB, it has the same effect as if they left too late and got
> caught in traffic that they normally would have avoided. Like I said, just
> two examples out of millions. -Dave

Come on - only 5% of their design capacity? Think about,
that is a ridiculous thought.

Ed

On 21 Jun 2005 13:20:51 GMT, Jim Yanik .> wrote:

>"JohnH" > wrote in news:FoudnbqsPaOIFCrfRVn-
:
>
>>> There's not a computer system in existence that could handle that
>>> task. Nor will there be.
>>
>> "Everything that can be invented has been invented."
>>
>> Charles H. Duell, Patent Commissioner, 1899
>>
>>
>>
>
>Please explain how the computers will know when a child runs out or a
>pedestrian is crossing the road,or Brent's unregistered bicycle is "taking
>the lane"?

Since when are bicycles or children allowed on freeways?