EMAIL ABOUT 'MERGING LANE' TRAFFIC JAMS
Q: What makes you such an expert?!
A: Nothing, I am not an expert. I'm just a guy who is fascinated by the
"traffic physics" I see during my commutes. I have no reputation to
defend, and I have no conviction that I'm a highly educated traffic
expert. Therefore I have no inhibitions about posting my observations and
reasoning. I hope it's revolutionary, but it might also have lots of flaws.
All the material on this website is nothing but speculation and "crazy
experiments" performed by a total amateur. For info from actual experts,
try some of these links.
Date: Wed, 15 Oct 1998
From: Scott Washburn
Subject: Entry to Guestbook
I just stumbled across this web site, and I must admit, it evoked a number
of reactions. First of all, it confirmed what those of us in the business
have long contended--that anybody with a driver's license thinks they are
a traffic expert. However, I was delighted to see a web site that provided
a forum for people to discuss their opinions about traffic and their own
personal driving behaviors. Unfortunately, though, I was disappointed to
see a web site that was purporting to educate people on traffic flow
behavior be based strictly on personal "theories" and absurd personal
"experiments". This web site contains a lot of conjecture and opinions
about traffic flow and ways to improve it. But, it is very obviously
missing solid mathematical and engineering based theories on traffic flow.
Now don't get me wrong, I do appreciate a forum for discussion on a topic
that impacts many people's lives everyday. However, I am concerned that
people that might not know better, which, unfortunately is a lot of the
people that have replied to this site, are led to believe that your
proposed driving behaviors are what should be adopted by everyone in order
to eliminate or minimize traffic congestion. Your personal experiments
with speed regulation and headway regulation make for interesting stories,
but that is about all. To think that your driving behavior alone in these
"experiments" had a significant impact upon the overall level of
congestion or the travel time experienced by commuters in this area is
ridiculous. Now, it is true that if you could get every driver on the road
to employ the exact same rational driving model, traffic congestion would
be lessened. But there is the problem, and where your analogy to fluid
flow breaks down (which has been explored by several researchers in the
past)--every water molecule (i.e., driver) in this case has a different
brain, and thus, every driver employs a different driving model. While it
would be nice to homogenize driving behaviour (e.g., a surrogate method
like vehicle computer control), and thus realize a system optimal traffic
flow model, the fact is that people act in their own best self interest,
so even the "good samaritan" actions of a few in the traffic stream will
be far overwhelmed by the individual self interest actions of the many.
Anyway, the fundamental problem with traffic congestion is that there are
too many vehicles trying to use the available capacity of the roadway at
the same time. And there is very little to nothing your individual games
of headway and speed regulation can do to prevent it. As I just mentioned,
roadway CAPACITY is one of the fundamental concepts that needs to be
understood...and I did not see that mentioned anywhere in your
"monograph". When demand exceeds capacity, congestion occurs-- it's as
simple as that. Thus, your hypothesis of the wall of state patrol cars
pacesetting the traffic is totally absurd. Trust me, this will do nothing
more than waste taxpayers' money. Now just think about it, since every
possible entry onto the freeway is not controlled (e.g., metering system),
it is not possible to prevent demand from exceeding capacity at any
particular point along that freeway. Thus congestion could occur anywhere
along the freeway where demand exceeds capacity, regardless of where the
wall of patrol cars are and what speed they are going. As the demand
increases in a freeway section (through increased on-ramp volumes, less
smaller off-ramp volumes), people will naturally slow down because of
their increased discomfort with decreasing headways as a result of the
increasing demand filling in the available capacity of the roadway. Now, I
will acknowledge that your attempt to explain a generally complex
phenomenon (i.e., shock wave theory) to the lay person is admirable, but
you should make sure you thoroughly understand it before trying explain
it. Unfortunately this is not the case, as is evidenced by your trying to
base your crazy personal traffic flow "theories" upon it. I apoligize if
this commentary sounds critically harsh, but I just had to shed some
factual light on this subject, based on real mathematical and engineering
research. Let me just close by saying this--drive in a manner that you
deem rational and leave it at that; no matter how many of your friends you
get to drive like you, it just won't make a difference when it comes to
changing levels of traffic flow and congestion.
Sincerely, Scott Washburn
Transportation Engineer
[
[ Note that you've missed an important point: my website is about my
[ empirical observations about the ability of ONE SINGLE DRIVER,
[ myself, to have a massive effect. It's not about theory, and I
[ need no friends to help me bust jams. If my article is correct,
[ then the efforts of people like yourself will not be able to stop
[ these ideas from taking over your field. It took a few years, but
[ they've started already:
[
[
[ Sept 1999
[ Nagatani shows that one driver with variable speed can trigger jams
[ http://prola.aps.org/abstract/PRE/v61/i4/p3534_1
[
[ July 2004
[ C. Davis finds traffic jams eliminated by unexpectedly small % of
[ cars automatically maintaining safe forward spaces.
[ http://link.aps.org/abstract/PRE/v69/e066110
[
[ To everyone else: never forget that experiment trumps theory. Since
[ I personally have erased traffic jams at merge zones in Seattle, then
[ no amount of theorizing by traffic engineers can change this fact.
[ (But perhaps I need some time lapse videos so you can see me do this.)
[
[ Also never forget that science often involves the "Emperor's new
[ clothes effect," where even a small ignorant child can point out a
[ stunning fact to which even the most highly educated experts were
[ hoplessly blind. (If the fact should trigger the experts' anger,
[ then we have a clue to WHY they were blind!) When it concerns
[ our learning to understand all the amazing and subtle features of
[ traffic-jam physics, perhaps it's better to be an ignorant
[ outsider. -billb
[
Date: Thu, 9 Jul 1998 01:11:29 -0700 (PDT)
From: "Michael S. Foster"
Subject: Entry to Guestbook
Aw, c'mon Bill, say you're not serious with your "simple" cure for traffic
jams. Surely you just posted this in order to invite people to poke holes
in this preposterous piece of illogic. No doubt your head has become
muddled while sitting still, frustrated in that jam on the I-5. My guess
is that your scheme would result in traffic coming to a complete
standstill, but that everyone would be able to change lanes as often as he
wishes, as long as he willing to back up. Of course, this is just a
guess, but then so are the results of your proposed scenario. Unless, of
course, you have the results of an experiment where you have convinced
thousands of people to behave in the manner you prescribe. Naturally, you
don't have such experimental results, because the only way to modify the
driving habits of this many people would be through the use of somewhat
draconian coercion. One must always be suspicious of simple solutions to
complex problems of human behavior, especially when it involves
surrendering rational self-interest to "altruistic" methods whose results
are doubtful. In reality, what we have here is a simplified model of
capitalism versus socialism and I think the experimental results ar e
fairly conclusive when it comes to these two systems. I therefore suggest
you put the atruistic traffic on the left and the self-interest traffic on
the right on your animated web page. Actually, there are a few
experimentally proven methods of reducing traffic jams. I travel a
certain section of the 405 freeway, sometimes several times a day, and I
noticed a rather dramatic reduction in traffic jamming when they (are you
waiting?) RAIS ED THE SPEED LIMIT! Of course, the speed limit was only
increased to 65mph from 55, but the difference was and is quite
noticeable. Dare I suggest that raising the speed limit to 75mph would
really have a salutory effect? Now, I realize that raising speed limits
isn't politically correct and won't give you that warm fuzzy altruistic
socialist feeling, but it might actually work. This particular rant
aside, Science Hobbyist is my very favorite web site and is an example of
the internet at its best. I have spent countless hours reading
practically everything here and following the endless trail of links
leading away to an incredible array of knowledge and information
previously unavailable to most people.
THANKS, BILL!
Michael S. Foster
Los Angeles, CA USA - Thursday, July 09, 1998 at 01:11:28 (PDT)
------------------------------------------------------
Date: Thu, 9 Jul 1998 09:39:45 -0700 (PDT)
From: William Beaty <
>
To: "Michael S. Foster"
Subject: Re: Entry to Guestbook: traffic
On Thu, 9 Jul 1998, Michael S. Foster wrote:
> often as he wishes, as long as he willing to back up. Of course, this
> is just a guess, but then so are the results of your proposed scenario.
> Unless, of course, you have the results of an experiment where you have
> convinced thousands of people to behave in the manner you prescribe.
Guess what... I DO have experimental results. Not thousands of people,
just one. It turns out that since traffic jams are single-lane events
once they appear, therefor sometimes A SINGLE DRIVER can have massive
effects on them. I personally have (temporarily) erased small and
medium-sized traffic jams at merging-lane areas on numerous occasions. If
EVERYONE maintained an extra space ahead of their car, then the
merging-lane traffic-jam dynamics would be seriously fouled up, and they
could not form.
I see one problem in your reasoning. What CAUSES the jam? Is there one
car far in the distance which is stopped, and therefor all cars behind it
must stop? Nope. These traffic jams are a form of turbulence. Tiny
actions by large collections of drivers cause them to appear. If the
collections of drivers behave differently, then traffic jams will be
affected. I've talked to one traffic dynamicist (see his link on my
TRAFFIC WAVES page) and he confirms some of my suspicions.
If we can discover which behaviors cause traffic jams, then we can
eliminate them simply by informing people about how to take personal
action. No legislation is needed.
> Naturally, you don't have such experimental results, because the only
> way to modify the driving habits of this many people would be through
> the use of somewhat draconian coercion.
Or by an ad campaign, as was done with recycling and energy conservation
in the last 20 years.
> One must always be suspicious
> of simple solutions to complex problems of human behaviour, especially
> when it involves surrendering rational self-interest to "altruistic"
> methods whose results are doubtful.
Why the hostility?
> In reality, what we have here is a
> simplified model of capitalism versus socialism and I think the
> experimental results are fairly conclusive when it comes to these two
> systems.
No, all we have is what we have: a suggestion that some types of traffic
jams will be affected by small changes in driver behavior. I'm convinced
that it only takes a small percentage of drivers to change behavior.
I see that something is not clear on my current webpage: it's not
necessary for ALL drivers to adopt the "altruistic" style during
merging-lane traffic jams. I've found that if only one driver does so (but
maintains a 20-car-length space), then a jam can be reduced or sometimes
even eliminated. Note that the "altruistic" style is not needed normally,
it's main effect is to alter the traffic-jam physics during heavy traffic
conditions where lanes of traffic are merging.
After many years of using my car to poke at commuter traffic and
considering the results, I have become convinced that only a small
proportion of drivers need to adopt the "altruistic" style during a jam,
yet it will have profound effects. You're right that no experiment has
been done such as my page depicts, with 100% of drivers behaving in the
abnormal way shown in the right-hand "gear-teeth merge" animatino. The
actual experiment used one car out of about 100: me.
> I therefore suggest you put the atruistic traffic on the left
> and the self-interest traffic on the right on your animated web page.
Why does this need to have ANYTHING to do with politics? Would you refuse
to experiment with traffic-holes yourself since traffic-holes are
obviously a despicable Socialist invention, while traffic jams are good
American Capitalism and we must protect them against the Commies? :)
Why instantly involve politics before even trying any experiments? But
since you have lept to a political stance, this means that I should go and
change my page to eliminate "altruist" and "self-interest", since it will
probably make many others leap into political-think and so refuse to
actually TRY my suggestions.
> Actually, there are a few experimentally proven methods of reducing
> traffic jams. I travel a certain section of the 405 freeway, sometimes
> several times a day, and I noticed a rather dramatic reduction in
> traffic jamming when they (are you waiting?) RAISED THE SPEED LIMIT! Of
> course, the speed limit was only increased to 65mph from 55, but the
> difference was and is quite noticeable. Dare I suggest that raising the
> speed limit to 75mph would really have a salutory effect?
It may. Since traffic waves and stoppages are caused by collective
behavior, then even tiny changes in the average behavior may have drastic
effects on them.
> Now, I
> realize that raising speed limits isn't politically correct and won't
> give you that warm fuzzy altruistic socialist feeling, but it might
> actually work.
Altruism is socialist? Only if it is forced by legislation. Actually, I
think there should be no speed limits at all. Look at Europe and the
Autobahn, it is experimental proof that such a thing is feasible. It has
always amazed me that the "land of the free" has no Autobahn. We are not
nearly as free as we tell ourselves.
If you can figure out a way to see traffic-holes as capitalistic (hey:
tiny individuals who can beat the hell out of a collective?), then give
them a try yourself. I've cancelled enormous numbers of "traffic waves"
on 520 during rush hour. (Check out the email discussion on my TRAFFIC
WAVES site.)
Anyway, glad you like my stuff!
((((((((((((((((((((( ( ( ( ( (O) ) ) ) ) )))))))))))))))))))))
William J. Beaty SCIENCE HOBBYIST website
amasci.com
EE/programmer/sci-exhibits science projects, tesla, weird science
Seattle, WA freenrg-L taoshum-L vortex-L webhead-L
Subject: Re: TRAFFIC JAM CURE
From: mac@mac.net (One Ton)
Date: 1998/07/08
Newsgroups: seattle.general
Buy a car with a rear view mirror and you might see things
differently....... ;>)
Mike
Subject: Re: TRAFFIC JAM CURE
From:
Date: 1998/07/09
Newsgroups: seattle.general
> Buy a car with a rear view mirror and you might see things
> differently....... ;>)
Hmmm? What's your meaning? In my traffic diagrams, wide gaps allow a
high-speed merge, while narrow gaps prevent merging and cause a traffic
jam. If this idea is wrong, I'd like to hear the details.
Actually, all attempts at "amateur traffic engineering" must be based on
the fact that drivers have very little effect on the traffic ahead of
them, while they can exert significant control on the traffic which
follows behind. If I were to "rubberneck" and completely halt my car for
no reason, I could leave a long-term traffic jam behind me. On the other
hand, if I approach a traffic jam at a slow speed, a large gap will open
up ahead of me. The gap allows the traffic jam to trickle away but with
no new cars piling up behind it, and when I finally arrive, part of the
jam has been "eaten" and converted into a region of slower traffic behind
me. In this way a single driver can take a bite out of a traffic
stoppage. If the stoppage was small, then it might be erased entirely,
and converted into a wide region of slightly slower traffic. Average
speed does not stay the same: The stoppage was a nonlinear effect, and
removing it causes increased overall speed.
But the "road rage" people will punish me! I'd better buy armament!
((((((((((((((((((((( ( ( ( ( (O) ) ) ) ) )))))))))))))))))))))
William J. Beaty SCIENCE HOBBYIST website
http://amasci.com
EE/programmer/sci-exhibits science projects, tesla, weird science
Seattle, WA freenrg-L taoshum-L vortex-L webhead-L
From: bitbucket@wolfenet.com (Sonya)
To: bbeaty@microscan.com
Subject: Re: TRAFFIC JAM CURE
Newsgroups: seattle.general
In article <6o36e2$3ll$1@nnrp1.dejanews.com>, bbeaty@microscan.com wrote:
> Actually, all attempts at "amateur traffic engineering" must be based on
> the fact that drivers have very little effect on the traffic ahead of
> them, yet they can exert significant control on the traffic which
> follows behind. If I were to "rubberneck" and completely halt my car
> for no reason, I could leave a long-term traffic jam behind me. On the
> other hand, if I approach a traffic jam at a slow speed, then a large
> gap will open up ahead of me. The gap allows the traffic jam to trickle
> away but with no new cars piling up behind it, and when I finally
> arrive, part of the jam has been "eaten" and converted into a region of
> slower traffic behind me. In this way a single driver can take a bite
> out of a traffic stoppage. If the stoppage was small, then it might be
> erased entirely, and converted into a wide region of slightly slower
> traffic. Average traffic speed does not stay the same: The stoppage was
> a nonlinear effect, and removing it causes increased overall speed.
This absolutely DOES work. In fact, I think you should buy the domain name
"www.trafficwaves.com" and I will personally volunteer to create and post
signs with the URL on every major freeway in the state.
Since reading one of the earlier posts about traffic waves, I've been
trying it out in traffic, and it's amazing. True, the hardest part is
ignoring the irritation of drivers behind you, but I just tilt up my rear
view mirror and think of something else. They seem to think that the two
feet they gain by being right on your tail will get them there faster.
Please keep posting this now and then - it should only take a small
percentage of drivers to noticably change the traffic situation in this
state.
Sonya
Please change "bitbucket"
to "elysium" in my address
to reply by e-mail.
Subject: Re: TRAFFIC JAM CURE
From: joearmy@wolfenet.com (Steven Spencer)
Date: 1998/07/14
Newsgroups: seattle.general
I've been doing the same thing and I've got to agree with the previous two
posters. This method of 'curing' traffic waves is amazing!! And Sonya,
just turn your music up extra loud to drown out the sounds of the idiots
who tailgate you when the honk. That's what I do!
And even more incredible is that my stress level while driving has dropped
tremendously. A large part of my frustrations have been the
rubber-neckers. Now I just don't worry about 'em as much.
Date: Fri, 17 Jul 1998 09:14:43 -0500
From: Bob Liepa
To: bbeaty@microscan.com
Subject: Traffic Merges
In cases where the number of lanes decreases, it is incorrect to state that
drivers who wait until the last second to merge are "greedy". It is also
incorrect to state that drivers who merge early show foresight and
consideration.
The optimal strategy for the entire group of drivers is to keep all the
lanes filled and to merge at the single point where the disappearing lane
actually ends. In fact, anyone who merges "early" is just contributing to
the problem.
There are probably two main reasons for merging early. One is that drivers
(mistakenly) think they are being courteous or socially responsible. The
other is that some drivers feel that if they wait to long to merge, they
might not be let in. In fact, it's always possible to merge.
If the disappearing lane is empty, one should therefore have no
reservations about using it until the end.
This is different, however, from "jumping the queue" by using an empty
acceleration lane from an entry ramp. In that case, using that lane would
require a driver to "unmerge" from traffic and then "re-merge" at a later
time. This creates an extra merge and is therefore not optimal for the
entire group of drivers. Of course, if the driver is legitimately entering
the highway at that point, he or she should take advantage of that lane
until it ends. It's a one-time privilege for newcomers to the traffic jam.
Bob Liepa
Senior Vice President
Research Dimensions Ltd.
30 Soudan Avenue, 6th Floor
Toronto, Ontario
M4S 1V6
Tel: 416-486-6161
Fax: 416-486-6162
Date: Sat, 7 Aug 1999 11:25:08 -0700 (PDT)
From: William Beaty <>
To: Chad Spencer
Subject: Re: Entry to Traffic Waves
On Sat, 7 Aug 1999, Chad Spencer wrote:
> Well you've solved the traffic problem in front of you, but you've
> totally neglected the situation behind you. Consider this: You're on
> the freeway, and you spot a traffic jam a mile or so up the road. You
> slow down to create the huge space in front of you, and by the time you
> get to the jam it is gone. BUT, while you were slowing down the traffic
> behind you, more cars were entering from the on-ramps at a normal rate.
You bring up a good point. I mention this problem in a couple of places,
but I guess I have to state the solution more specifically.
Yes, it is impossible for me to remove cars from the road. All I can do
is transfer the patches of high-density traffic from one place to another.
I can also take a small, dense region and spread it out into a large,
less-dense region. In a highway with no entrances or exits, wouldn't this
be pointless? No, not always, because I can erase "traffic waves" and
"shock wave" stoppages. A traffic wave is nothing but patches of
slowdown, and by moving them around, I can create totally smooth traffic.
And a total stoppage can be changed into a large zone of slow traffic.
There is a second place where I can have an impact. When a slowdown-zone
becomes stuck at a merge-zone, a true bottleneck appears. My little
animations depict this phenomeonon. If I can grab that slowdown and move
it far upstream, so that cars at the merge zone can merge at high speed
like "gear teeth" or a "zipper", then I have changed the capacity of the
highway at that point. It is a nonlinear effect. If it wasn't, then
there would be no good reason to mess with traffic there.
Is it possible to make traffic worse? Certainly. I don't entirely
understand your message, but I think you might be discussing a situation
where there are a number of merge-zone traffic jams in a row. If I try to
"cure" the last one in line (the most-forward downstream jam), then I make
the jam worse at the upstream ones. Yes, this is correct. If most of the
traffic is coming from entrance ramps, and not from the main highway, then
one car can safely manipulate only the first jam, the most-upstream jam.
If I try to cure the downstream ones, then I dump a slowdown into the
upstream ones and make them worse (or possible trigger a new traffic jam.)
Another way to say it: a row of small traffic jams will interact with
each other. Each one is a nonlinear "switchable" phenomenon all on its
own. Together, they are like a living thing and their actions will be
very unpredictable. If I try to manipulate them, they might "explode"
into a really horrible mess.
In my commutes I've encountered big jams at single exit ramps, where there
are no other exit ramps anywhere around. I fortunately don't drive in a
place where those "living creature" collective multi-ramp jams exist, so
I've never had to think about them.
> I'll give you a big hand for effort, but in all of the thought you put
> into eliminating the traffic in front of you, you completely forgot
> about what was going on behind you.
Of course not.
It is impossible for me to affect the traffic ahead of me. All my
thinking is based on this fact. I can only affect the traffic behind me,
so that's the first thing I think about. See
http://trafficwaves.org/seatraf.html#three How can I
manipulate the traffic behind me? Since I cannot drive faster in
congested traffic, I can only drive slower. If I drive slower INSIDE a
congested zone, then I make the traffic worse. Therefor I must create my
"antitraffic" hole while I'm in a lightly congested zone, and not do this
in the middle of a bunch of on-ramp merge zones!
((((((((((((((((((((( ( ( ( ( (O) ) ) ) ) )))))))))))))))))))))
William J. Beaty SCIENCE HOBBYIST website
http://amasci.com
EE/programmer/sci-exhibits science projects, tesla, weird science
Seattle, WA freenrg-L taoshum-L vortex-L webhead-L