typodupeerror

## Math Says You're Driving Wrong and It's Slowing Us All Down (wired.com) 404

A new study in IEEE Transactions on Intelligent Transportation Systems mathematically suggests that if you and everyone else on the road kept an equal distance between the cars ahead and behind, traffic would move twice as quickly. From a report: Now sure, you're probably not going to convince everyone on the road to do that. Still, the finding could be a simple yet powerful way to optimize semi-autonomous cars long before the fully self-driving car of tomorrow arrives. Traffic is perhaps the world's most infuriating example of what's known as an emergent property. Meaning, lots of individual things forming together to create something more complex. Emergent properties are usually quite astounding. You've probably seen video of starlings forming a murmuration, a great shifting blob of thousands upon thousands of birds. Bats flying en masse out of a cave is another example, swarming sometimes by the millions through a small exit. And scientists are just beginning to understand how they do so.
This discussion has been archived. No new comments can be posted.

## Math Says You're Driving Wrong and It's Slowing Us All Down

• #### Merge problem (Score:2, Insightful)

if you and everyone else on the road kept an equal distance between the cars ahead and behind, traffic would move twice as quickly.

Yes, because no one would be merging into traffic anymore.

• #### Re:Merge problem (Score:5, Insightful)

on Saturday December 30, 2017 @11:53AM (#55834235)

if you and everyone else on the road kept an equal distance between the cars ahead and behind, traffic would move twice as quickly.

Yes, because no one would be merging into traffic anymore.

If everyone kept an equal distance and followed a standard merging pattern of every other car, then it would likely solve the merging problem as well. Long ago we were shuffling decks of cards in a much less practical and inefficient way until certain physical moves were found to increase that efficiency ten-fold.

It's also well-known that impatience creates stop-and-go traffic patterns, which is but one of the many human factors that autonomous solutions will be looking to solve.

• #### Re: (Score:2)

If everyone kept an equal distance and followed a standard merging pattern of every other car, then it would likely solve the merging problem as well.

No, because the distance between the cars before the merge point would be bigger than the distance after the merge point, which contradicts the assumption that all distances would be equal.

• #### This is why (Score:4, Funny)

on Saturday December 30, 2017 @02:19PM (#55835091)

This is why I always pull into your lane when you try to pass me. You are screwing up the algorithm and I'm fixing it.

• #### Re:Merge problem (Score:4, Informative)

on Saturday December 30, 2017 @01:16PM (#55834721)
Exactly. Overall traffic flow seems most efficient if traffic on main routes is bunched into "packets" of vehicles with gaps between packets large (long) enough to allow traffic from smaller routes to enter or merge onto the main route. My home town did this with updated lights and sensors about thirty years ago (!). Once on the main road you never had to be stuck at a light until all reaching the next town.
• #### Traffic in Mumbai (Score:3)

on Saturday December 30, 2017 @11:40AM (#55834177)

I was trying to get from the Domestic to the International terminal, about 5km, in a cab (along with wife and luggage). The lightrail system was under construction, which added to the mess. Traffic was gridlocked until about midnight, when it started moving. Turns out there was one traffic light that caused the gridlock. Once that went to flashing yellow, the drivers negotiated their way through the intersection.

• #### No clickbait headlines (Score:2, Insightful)

Math says you're treating Slashdot readers wrong and it's making the internet worse for all of us.

• #### Re:No clickbait headlines (Score:5, Informative)

on Saturday December 30, 2017 @11:58AM (#55834267)
Common sense says we don't need math to show what is slowing down traffic. Maintaining distance really means matching speed, acceleration, and deceleration. If those happen distance itself becomes less of a key factor. Added distance is just an aid for drivers who aren't good at staying with traffic flow.

If you want traffic improvement,
1) get left lane laggards to drive properly and not slow down faster traffic
2) get everyone to be expeditious when intersection lights turn green
3) teach people not to contribute to traffic compression waves by over decelerating and then under accelerating
• #### Re: (Score:2)

It's hard not to be a laggard, and always expeditious, while simultaneously not over accelerate/decelerate.

Common sense says we don't need math to show what is slowing down traffic.

Slow traffic is caused by trying to put more cars on a stretch of road than it can handle.

• #### Re: (Score:3)

Slow traffic is caused by trying to put more cars on a stretch of road than it can handle.

True in many but not all cases. If you increase throughput by managing congestion, you reduce the number of cars on the road at a given time.

• #### Re: (Score:3)

How am I supposed to do all that shit when I've committed 96% of my attention to my iPhone?
• #### Re: (Score:2)

The reason math won is that you don't need to dampen spike if nobody creates spike.Theory vs reality.

• #### Re: (Score:2)

The funny thing is that the 'left lane laggards' that cause the problem are quite often speeders. I see this all the time. Someone in the right lane wants to pass, so they move into the left lane and start the pass. They are moving faster than the right lane, probably faster then the speed limit. They are not a 'laggard'. But then some aggressive speeder can't be bothered to notice (or care) that traffic is going slower, so he comes right up on the car in front and BRAKES. Now he is going slower than

• #### Re: (Score:2)

TLDR

Impression: You think there's nothing wrong with camping in the passing lane going 55.0001 MPH.

• #### Re: (Score:2)

I neither said nor implied any such thing. If you are going 70MPH passing someone going 65, a properly executed pass takes about 30 seconds. It is entirely reasonable for someone going 70 to pass someone going 65.

During that 30 seconds, someone doing 80 will travel about 500 feet more than the person doing 70. If the person doing 80 was 500 feet behind the person doing 70 when the pass started, they will encounter each other at the very end of the pass for a very short time. At any time during that 30 s

• #### Re: (Score:2)

yep, this. unfortunately, this would require at least 70% of the people driving to divert an additional 15% of their interest towards the activity of driving. how do you accomplish that?

• #### Re: (Score:3)

Meh, I don't think it'll get much better than today. Commute driving is a solid mix of people who are either:

a) Late, stressed and glued to the rear bumper in front of them
b) Bored, zoned out and mentally passing the time with something else

It's human nature that you'll have a huge variation in reaction time and aggressiveness to close the gaps. I have a good view of that near work, due to a slope, bridge and intersection on the other side you can see probably 30-40 cars at once at rush hour. You see the gr

• #### Re: (Score:2)

If you want traffic improvement, 1) get left lane laggards to drive properly and not slow down faster traffic 2) get everyone to be expeditious when intersection lights turn green 3) teach people not to contribute to traffic compression waves by over decelerating and then under accelerating

"Get other people to do stuff" is not usually a productive strategy ...

What you do have the power to change is your own driving [trafficwaves.org].

• #### Re: (Score:2)

lol i remember that site from like 1998. can't believe it's still around.

• #### left lane laggards (Score:2)

>1) get left lane laggards to drive properly and not slow down faster traffic

With new legislation effective las October 1, the Nevada Highway Patrol is now issuing "obstruction of traffic" tickets to cars in the left lane that "know or reasonably should know" that their slow speed is obstructing traffic.

hawk

• #### Nothing more annoying (Score:3)

on Saturday December 30, 2017 @11:50AM (#55834221) Homepage
than getting stuck behind a driver that keeps racing up to the car in front of them and then hitting the brakes, falling back and doing it all over again. Keep it steady man.
• #### Re: (Score:3)

It's incredible how many people can't even maintain a constant speed on a motorway/highway. If you turn on cruise control you quickly find that people randomly accelerate and decelerate.

I also find that when trying to overtake people they often speed up. I think it's unconscious, at least I hope it is because otherwise it's a really stupid thing to do.

• #### Re: (Score:3)

It's incredible how many people can't even maintain a constant speed on a motorway/highway. If you turn on cruise control you quickly find that people randomly accelerate and decelerate.

Depends where.

In NM for example, once you set the cruise control, you do NOT touch the cruise control at any time for any reason. Its forbidden. Also, you don't use the left lane because that's for slow people. And you don't use the right lane because that's for fast people.

The result is that you'll be driving along and over

• #### Lane for automated vehicles (Score:2)

In the coming years, there will no doubt be 1-2 lanes on federal highways just for automated driving. Ideally, it will have variable speeds all the way up to 100 on good days. Keep in mind that federal highways were designed for 120 mph minimum.
• #### Re: (Score:2)

Except some redneck in a jacked up Toyota pickup will always drive in the designated lane.

• #### bumper to bumper (Score:2)

When I drive bumper to bumper in traffic ( equal distance between the cars ahead and behind) I move at a crawling speed; If people increased that space we would move twice as slow? I did not think that was possible.
• #### In other words, there's an optimal distance. (Score:5, Insightful)

on Saturday December 30, 2017 @12:01PM (#55834289) Homepage Journal

That's not surprising. Spread cars out too much and you reduce the roadway's capacity. Put them too close together and you have to slow down to accommodate the driver's minimal reaction time. Having every driver choose his own distance means you can have both effects simultaneously: wasted space and insufficient response time.

Put all these constraints on and it seems obvious that you want to space cars uniformly with the minimal distance consistent with whatever statistical level of safety you demand. Naturally robotic systems will be more efficient since they require less response time -- in fact they can react to events that will cause the car in front to slow.

What would be interesting is to see the exact results they came up with: how far for how fast and under what conditions? What are the significant input parameters of the model? For example I'm sure varying the acceptable probability of a crash has a powerful effect on the optimal distance.

• #### Re: (Score:2)

Put all these constraints on and it seems obvious that you want to space cars uniformly with the minimal distance consistent with whatever statistical level of safety you demand

If all the cars are going at maximum speed with minimum safe distance between them, it becomes impossible for a single car to merge.

• #### Re: (Score:2)

That's why I'm interested in the actual details of the model. Exactly what you are optimizing and how you describe it makes a big difference.

For example widening a highway clearly increases its capacity, but if the capacity of the roads it feeds is limited you just end up turning a long skinny traffic jam into a short fat one. I've seen this happen on multiple occasions. The parameter being optimized (throughput) was the wrong one.

• #### Re: (Score:2)

Just shove your way in, don't telegraph your intent by signaling - that just lets them know to cut you off. Ohhh, you meant safely.
• #### Re: (Score:2)

What would be far more interesting is to teach people the effective methods of driving as part of a test for competency. Here in the USA it's very minimal, focused on regurgitating rules mostly no one obeys. How about looking 2-3 cars ahead and preparing to brake/accelerate ahead of time (feed forward control)? How about speeding up to merge instead of slowing down? How about actually zipper merging instead of some genital posturing contest? There are a dozen things that reduce or eliminate congestio
• #### Re: (Score:2)

You mean slamming on the breaks and cutting the wheel to try to get off the road isn't the thing to do when you don't have traction?

• #### Re:In other words, there's an optimal distance. (Score:4, Insightful)

on Saturday December 30, 2017 @01:01PM (#55834637)

I always try to have the distance to the car ahead of me set so I never have to hit the brakes and when the traffic does slow up, not having more than the usual 2 second gap just before it speeds up again. I feel like this results in me maintaining the highest speed possible. Of course people usually cut in front of me, so I have to slow down more than I would otherwise.

This is only for freeway traffic, city streets and inconsistent stoplights are a whole other ballgame.

• #### Re: (Score:2)

A column of traffic can act like a one dimensional fluid. If everyone drove as you did, that fluid would be compressible -- any local variation in speed would tend to propagate slowly and continuously, if at all. But if everyone tailgates, the column of traffic acts like an *incompressible* fluid. That means when you tap the brakes, it generates something analogous to a shockwave which can propagate faster than the traffic itself is moving. In fact that's the norm in heavy traffic.

• #### No there's no optimal distance (Score:2)

The way TFA it explains it is pretty silly. Keeping the same distance to the car behind you just means the car behind you is keeping the same distance from you as you are from the car in front of you. i.e. the cars are all equally spaced. Things like one car tapping its brakes, failing to maintain speed, or merging into the lane causes the cars behind it to slow down. If the cars behind this one car are bunched up, it exacerbates the slowdown. The optimal way to space the cars to minimize this slowdown
• #### Re: No there's no optimal distance (Score:2)

They are not saying equal distance across all units like a train. Trains do not have units coming and going. Nor do detached units have a central authority setting their individual speeds.

You can have units at ...1,5,9,13... distance (there is no beginning nor end). And if a unit mergers after #1, it can become 1,3,5,9,13 => 1,3,6,9,13 => 1,3.5,6,10,13 => 1,3.5,6.25,10,13.5..., etc. (of course 1 would move forward too but did not for simplicitys sake).

With units coming and going, there would rare

• #### Re: (Score:2)

a block of a dozen or so cars on cruise control radar-locked bumper to bumper - with larger spaces in front of and behind the train where new cars can merge into.

That only works as long as there are larger spaces. Too many people try to merge, and the larger spaces get smaller, until you reach the point where you have to deny traffic from the on ramp.

• #### And stop competing for money, too (Score:2)

... because obviously, if maximizing the efficiency of human work in terms of how many products/services are created is the defined goal, it makes absolutely no sense to have a multitude of competing companies spend lots of effort and material into developing/producing/advertising the same kind of product/service.

The only catch to this, just as with human car drivers: Not compatible with homo sapiens, which evolution shaped over millions of years to behave competetive and give a shit about some "greater g
• #### Can't drive faster than... (Score:2)

... speed limits.

Artificially low speed limits.

• #### Long been taught in driver's education (Score:2)

Seriously? It takes the IEEE to tell us of something which has been known since before "The Godfather" hit the screens??

• #### Remember folks... (Score:2)

Never drink and derive.
• #### Set adaptive cruise control (Score:5, Insightful)

on Saturday December 30, 2017 @12:44PM (#55834531)

Just set your Tesla (or other modern) adaptive cruise control to, say, five car lengths, and just steer. It is far far easier than having to brake/accelerate and the hardware watches even when the driver has zoned out. No worries about hitting the idiot in front. No worries if someone merges into your lane: the car adapts.

• #### Re: (Score:2)

Oh cool you're going to buy me a "Tesla (or other modern) adaptive" vehicle?
Not everybody has that luxury.

• #### M25 variable speed limit. (Score:2)

In the UK, the M25 (circular road around London) has a variable speed limit.

As traffic becomes heavier, the speed limit drops. But all lanes get the same speed limit. The limit is heavily enforced with cameras.

With the limit set quite low, the traffic proceeds much more uniformly, there is no advantage to changing lanes, everyone drives at the limit.

The result is that more cars can flow past any point on the road. It's another example of the effect predicted by the paper.

• #### drive much? (Score:2)

Nobody who drives *EVER* tries to monitor the distance between their car and the one BEHIND them! It's your responsibility to monitor the distance to the car in front of you. This is why all the seats in an automobile face FORWARD. Duh!
• #### Fixing Traffic (Score:2)

If you want to fix traffic in densly populated areas, you should avoid cars and use mass transit. Light rail, trams and modern busses can achieve higher travel speeds and higher throughput than cars. To connect less populated areas, you use park and ride systems, i.e. , parking lots with direct bus/tram/train access.

• #### No, I'm doing it right! (Score:2)

It's just the idiots around me that are doing it wrong!

• #### We'll do that (Score:2)

> you're probably not going to convince everyone on the road to do that
We'll do that when cars are driven automatically and AI does that, automatically.
• #### Who is this "you" you're talking to? (Score:2)

Math Says You're Driving Wrong and It's Slowing Us All Down

Fuck you and your condescending clickbait headlines. You don't know what I'm doing.

if you and everyone else on the road kept an equal distance between the cars ahead and behind, traffic would move twice as quickly. Now sure, you're probably not going to convince everyone on the road to do that.

Damn right you're not, because it's a fucking stupid idea. You want drivers to monitor the distance to the car behind them? There are enough problems getting drivers to concentrate on the direction they're travelling in.

• #### Forward March (Score:2)

by Anonymous Coward

Master of the obvious. Armies have known this for millennia. That's why soldiers maintain fixed distance and all start out together with a "forward...march" command and stop with a "company...halt". Try that at the next traffic light when you are 5 cars back. Man, if everyone just all went at the same time when the light turned green it would be awesome, but without an automated system it's just a 4 car pile-up.

With human drivers you inevitably get the "slinky" effect due to reaction time and differences in

• #### Self-driving car transition (Score:3)

on Saturday December 30, 2017 @01:24PM (#55834755) Homepage

I believe this will eventually become the force driving the adoption of self-driving technology. When we get to the point autonomous cars do it right and are mixed in with human-driven cars that are screwing up and slowing down the traffic pattern for everyone else. I can see the current driving model being totally turned on its head with commuters eventually demanding that we ban human drivers.

There will also be economic pressure. Human drivers need signs, lights, and a weighty infrastructure investment. Autonomous cars need none of that expensive support.

• #### Bah! Really? (Score:2)

I was once in a traffic jam that had traffic backed up for miles. All the cars were equal distance apart, but none of them were moving! Turns out there was no accident ahead, no stalled cars, no reason whatsoever.

What happens is some idiot lightly taps his brakes, may to turn off his cruise control, who knows. Then the driver behind him, not knowing how hard the first one is braking, taps his brakes a bit more. This continues with drivers braking a little harder all the way back, so that cars a hundred yard
• #### Re: (Score:2)

Better driving results in a slightly higher road capacity, but at the cost of reducing the margin for correcting small mistakes. At some point, someone inevitably will make a slightly bigger mistake, and mess it all up. The closer you get to optimal road capacity, the worse the consequences will be.

• #### obligatory (Score:3)

on Saturday December 30, 2017 @02:17PM (#55835073)
This guy [trafficwaves.org] figured all that out in detail years ago ...
• #### Their math is wrong. (Score:2)

If you have a road containing x lanes and a flow rate of y, then you can optimize traffic based on the predictive analysis of the population and when they *screeech* OMG WTF IS THAT ASSHOLE DOING! And then the traffic flow is zero. Q.E.D

• #### CGP Gray's -- The Simple Solution to Traffic (Score:3)

on Saturday December 30, 2017 @02:51PM (#55835249)
• #### You can't fix a system by changing *everybody* (Score:5, Insightful)

on Saturday December 30, 2017 @02:57PM (#55835283) Homepage
I work in a manufacturing environment, and changing even a handful of people's behavior is so incredibly difficult and costly ("always pick up one orange nut at a time, then the blue nut, don't grab two at once.") that asking everyone to change the way they drive is just ridiculous. You have to change things so that the desired behavior is the easier behavior. For instance, advanced cruise control that adjusts your distance automatically might be a solution. In our plant, if the process says they should do X before Y, then the only way to ensure it actually happens all the time is to prevent Y from happening until there's proof X happened. People just aren't reliable.
• #### Same speed in same lane good, different lane bad. (Score:4, Informative)

on Saturday December 30, 2017 @04:06PM (#55835625) Journal

Keeping constant spacing and running at a reasonable speed within a lane may be good. But holding the same speed in adjacent same-direction lanes is very bad.

In driving classes, back in the mid-20th century, we were warned against it. You NEVER were to hold the same speed as a car in an adjacent lane. (About a 5 MPH drift, with leftward lanes faster, was close to ideal.) Judging by the behavior of current drivers on California freeways that lore has apparently been lost.

Some of the issues:
- Adjacent cars form a multi-lane "rolling roadblock". Drivers behind them who wish to travel faster are impeded, collect behind them, and end up "compressed", setting up the conditions for a chain, reaction multicar pileup.
- With an inter-lane drift a driver wishing to pass a slower car soon has an opening to switch lanes and proceed.
- With the slowest lane to the right and increasing speed to the left, merges and exits require less speed change and have better timing margins, long-distance traffic proceeds rapidly with little disturbance, and lane changes are easy. Drivers have the opportunity to rapidly distribute themselves among the lanes and drive at a speed where they're comfortable.
- When driving at the same speed as an adjacent vehicle you increase your risk of collision:
- If you're in a blind spot you STAY in the blind spot for a long time. The window of opportunity for the adjacent driver to happen to make a lane change into you - or into the space immediately in front of you, becomes much larger than if you had a relative drift.
- If you hold relative position the other driver's peripheral-vision motion detector doesn't keep him aware of your presence. After a minute or so you're likely to fall out of his attention. Then, if a sudden traffic situation makes him need to change lanes suddenly (or he just wants to change lanes and forgets to do a recheck), he may swerve into you.

(By the way: The two-way two-lane equivalent of the rolling road-block chain-reaction-collision precursor is the "rat pack", a term of art in traffic engineering. It occurs when the first driver goes slightly over the limit and the second driver won't pass because he doesn't want to risk the necessary speed, but follows too closely for following cars to pass in two single-car hops. Fault is primarily on the second driver.)

#### Related LinksTop of the: day, week, month.

The person who can smile when something goes wrong has thought of someone to blame it on.

Working...