The future of speeding

[CitizenKane]

Do you think speeding will be possible in the future? When all cars are part of a network, all data from the cars shared and cars are primarily self driving, will society accept the possibility of speeding? As a freedom lover I am not particularly thrilled about the idea of never being able to go faster than the legal speed limit, but in all honesty - it is a freedom I love to have myself but would not mind seeing other not having to increase the safety on the roads. Statistics are crystal clear - reducing speed dramatically reduces traffic fatalities. I think a lot of countries will opt to stop the possibility of speeding, weighing the interests of individual freedom against saving lives. This could be done either by technically preventing speeding or by fines being "autoissued" when you exceed speed limits, since all data of your driving will be available.

On the positive side, I still think network and self driving will mean faster transportation anyway and the vastly improved safety may even allow higher speed limits. Self driving cars in a network should theoretically be able to go 100 mph with only a few meters between them on the highway without any major safety concerns.

Human drivers wanting to speed may have to go to closed off racing circuits.

What do you think?

 

[brianman]

Statistics are crystal clear - reducing speed dramatically reduces traffic fatalities

Assertion -> Links please.

Also snarky response: It certainly is incorrect if I slam on my brakes with a car behind me on the freeway with no traffic in front.

 

[CitizenKane]

Assertion -> Links please.

Really? Where to start...

Do speed limits reduce the number of road deaths?

 

[jbcarioca]

...Statistics are crystal clear - reducing speed dramatically reduces traffic fatalities..."...

What do you think?

On the truth-o-meter test this statement gets a "half-correct". Any responsible actuary dealing with the subject knows that crash frequency rises with variation in vehicle speed, but not with speeds themselves. Vehicles traveling at the prevailing speeds on the road are least likely to have an accident. However, loss severity does increase with higher speeds. These propositions have many, many studies proving the point. They are true.

However, "speed kills" is much like global warming denial. If somebody notes that a day is the coldest ever recorded somewhere that is taken to show global warming is a fraud. Similarly, the data proving beyond any shadow of a doubt that traffic speed variability is the culprit most correlated to accident frequency is a bit too subtle for many bureaucrats, who find it easier to collect revenue from speeding tickets than to find out how to smooth traffic flows to reduce accidents. In the end low speed limits are about revenue collection.

I'll skip the links. A couple are already in the thread. The detailed data is generally not easy to find in public sources because vested interests are not anxious to hear about developments which, from their perspective might be hard to explain and might reduce revenue anyway. What politician wants that?

 

[Tam]

...Statistics are crystal clear - reducing speed dramatically reduces traffic fatalities...

Fastest Road in America: 85 MPH and We May Be Going Even Faster

"Don't be surprised if you hear growing calls for states to wave speed limits during certain periods. In the late 90's Montana had no speed limit during daytime hours....

Supporters of Montana not having a speed limit say the fatality rate on Montana highways in the final five months without a speed limit were lower than after speed limits were reinstated...."

Also:

 

[S4WRXTTCS]

I said yes because essentially speeding right now is something reserved for the rich.

It will simply remain that way.

 

[thegruf]

Almost invariably speed is equated with accidents.

We regularly get figures quoted that over half of accidents are caused by speed.

This is a factual inaccuracy. The data (in the UK) states that only appx 5% of accidents have their primary cause as excessive speed.
The largest cause of accidents is errors in observation.

It is however correct to state that the majority of accident involve cars that are travelling at excessive speed (even this is not necessarily news as most cars speed to some extent), which clearly has the potential to increase the severity of the accident, but it is not the priamry cause.

As technology improves it is inevitable that speed controls will become more "interactive".
We see this in particular in the UK already with "black boxes" being fitted to new drivers vehicles by insurance companies with the premiums varying according to driver behaviour. Early indications are that this works very well.

In a hypothetical world where all vehicles are autonomous clearly the driver will little if any control over the speed.

That's what trackdays are for

 

[S4WRXTTCS]

The problem with speeding is trying to define what speeding is.

Is it 5mph over the speed limit?
Is it 10mph over the speed limit?
Is it 15mph over the speed limit?

What classifies as speeding depends on where you're at.

My understanding is it's not speed that kills, but speed differential that kills. I prefer autonomous cars because it will keep everyone roughly at the same speed. Where people care more about time of arrival than the actual speed they go. They'll likely be able to pay more to have priority.

 

[CitizenKane]

On the truth-o-meter test this statement gets a "half-correct". Any responsible actuary dealing with the subject knows that crash frequency rises with variation in vehicle speed, but not with speeds themselves. Vehicles traveling at the prevailing speeds on the road are least likely to have an accident. However, loss severity does increase with higher speeds. These propositions have many, many studies proving the point. They are true.

However, "speed kills" is much like global warming denial. If somebody notes that a day is the coldest ever recorded somewhere that is taken to show global warming is a fraud. Similarly, the data proving beyond any shadow of a doubt that traffic speed variability is the culprit most correlated to accident frequency is a bit too subtle for many bureaucrats, who find it easier to collect revenue from speeding tickets than to find out how to smooth traffic flows to reduce accidents. In the end low speed limits are about revenue collection.

I'll skip the links. A couple are already in the thread. The detailed data is generally not easy to find in public sources because vested interests are not anxious to hear about developments which, from their perspective might be hard to explain and might reduce revenue anyway. What politician wants that?

I'd say its highly debatable which side of that argument most resembles climate change deniers. In both cases it seems to me some people ignore facts the consequences of which they don't appreciate. Saying politicians hide the truth to protect revenues (speeding tickets?) sounds very much like climate change denier conspiracy theory to me.

Anyhow, even if the premise that higher speed equals more traffic deaths would be wrong, it matters little for this question since an overwhelming majority of policy makers do believe speed kills.

 

[Nikxice]

Even in an autonomous driving world there are too many factors associated with speeding that would rule out blanket acceptance. Speeding on highways in Montana would probably only carry a slightly higher risk of a crash. Vehicles could still be safely staggered to compensate for potential problems. Speeding on urban highways is more complicated. The higher speeds would require more space, spreading a large number of vehicles out over a fixed amount of real estate. The theoretically driving "with only a few meters between them" would solve spacing issues, until you step back and imagine millions of vehicles trying to safely accomplish this on a daily basis. The inevitable mechanical failures, road debris, black ice, you name it, would likely weigh heavily against increasing speeds much anytime soon. In a hundred years, who knows?

 

[theboom1]

My shadow mode idea(or whatever you want to call it just the basic concept of it) could allow the people that like to speed to be able to.

I envision something with shadow mode that people that like sports cars would like even more than what we have now. If shadow mode or something similar makes it impossible to crash in injure someone, then when there is not a lot of traffic (monitored by the computer) then mr officer you wouldn't mind me and my buddy doing a red light race right in front of you right? Or me do a quick 0-150 pull? See where I'm going with this? It could be amazing. Of course we need to only allow cars that are designed to go that fast be able to. We don't need a civic going 150. We could use ludicrous mode whenever we wanted to its fullest potential and it be legal.

 

[adiggs]

To some extent, the definition of speeding is a function of poorly set speed limits for roads (as others have pointed out earlier, a revenue thing, but also a historical and inertia thing IMO). A road I find a joy to drive on is a toll road in Oklahoma - speed limit 75 (or 70 - can't remember specifically right now), with the additional proviso - "No Tolerance".

Two features - the published speed limit is actually appropriate for the road conditions and design, and from other vehicle's behavior, I've deduced that 1 over is probably pushing it. In that situation, when you pull onto the road and see a car most of a mile in front of you, that car is going to stay a mile in front of you. They don't get further ahead, and you don't catch up. Besides moving things along, the road has always seemed particularly safe to drive on to me. Of course, I don't have access to actual statistics from accident or death rates on that stretch of highway - I just know as a driver, I like having a clear and specific standard, that is both appropriate for the conditions / design, and is reasonable.


I love that. If a road is really a 65 mph road (as is common where I live), with a posted speed limit of 55, then that's a poorly chosen speed limit. It contributes to a wide spread belief and behavior that the real speed limit is +5 or +10 (or +8, or +13, or ...). Our road designs are improving over the decades, even if the maintenance isn't always keeping up, and cars today are better able to move safely at these higher speeds.

As autonomous driving enters the world, and cars gain the ability to queue up on the highway and all go exactly the same speed, we will lose a lot of the efficiency reasons for driving more slowly. Will our speed limits continue to be artificially low by 5-15 mph? Except THEN, the technology might enable +30, +45, or +60 driving.

 

[bxr140]

I'd say its highly debatable which side of that argument most resembles climate change deniers.

I was thinking the same exact thing, especially since that response doesn't really contradict your assertion. Yes, excessive speed is a problem and yes, the sub-bullet is that it's often speed delta and not absolute speed...but autonomy or not, nobody is going to argue that 140mph is as safe as 70mph. Citing studies that show elevated limits aren't more dangerous is an exercise in half truths.

There are a few absolutes that will bound the future of high speed travel:
1. Humans will always maintain a desire to push the limits, whether it's of themselves or their machines.
2. Going faster is exponentially inefficient.
3. Autonomous operation is safer than human operation

I believe for the most part human operated traffic finds a happy medium between safety and convenience (basically, speed) right now, at least for surface streets. While autonomy might allow for a little faster flow of traffic, we're not going to be adding 20-30mph to that current 25 or 35 zone the day we give Jesus the wheel.

For limited access roads between population centers I think there's often quite a bit of room for speeding up the flow of traffic, and that's where we'll see increases of 20-30 and even 40mph sometimes...but there will still be a cap. Autonomy or not there's still reaction time in the system for the random unexpectedness, and there's still the pesky range problem too. Gut feel is that those aspects combined will limit even fully autonomated traffic to somehin like 90-100mph on the open road.

 

[adiggs]

...

For limited access roads between population centers I think there's often quite a bit of room for speeding up the flow of traffic, and that's where we'll see increases of 20-30 and even 40mph sometimes...but there will still be a cap. Autonomy or not there's still reaction time in the system for the random unexpectedness, and there's still the pesky range problem too. Gut feel is that those aspects combined will limit even fully autonomated traffic to somehin like 90-100mph on the open road.

For the initial implementation, I have different reasons, but I get to the same conclusion - at least for the USA -- new speed limit of 90-100 for autonomous cars in communication with each other. Even that will come in steps due to the speed variance / safety issue (sharing the highway with merely human drivers). To offset the range penalty, the cars will follow each other very closely under machine control, so that they brake and speed up together.

I don't see that as the peak though. I see that as an initial implementation that is reasonably achievable and that will help us start figuring out what it means for independent objects to move rapidly in close formation.

Afterwards, with more and more traffic moving at these higher speeds, I foresee more development focused on vehicles being able to move at faster and faster sustained speeds. Though I've not been there myself, think of the autobahn users that find the inability to sustain 120 mph for 2 hours to be unacceptably slow, and of unacceptably short duration. I'm the first to agree that's a limited use case today, but now we're positing a new use case that will make this high speed and duration more common (namely, spreading it to more countries). More common will drive more investment, and more capable vehicles into the fleet.

And one solution for going fast for long periods of time, is to do it in a group of vehicles. Every vehicle gets better efficiency than being alone, even the vehicle in front of the string. If we've got that technology, then it shouldn't be all that difficult for the string to rotate the car at the head of the pack periodically, to extend that vehicles range and at least get it to it's next charging stop.

 

[Ardie]

We speed demons should be careful of what we wish for.

California's Department of Transportation did a study a few decades back on "what is the most efficient speed," which greatly impacts what cities will do when its time to spend gobs of money in an attempt to improve the Interstate Highway System threading its way through town.

The most efficient speed in the Greater Metro Area of Los Angeles is ... 30-35 mph.

This is because the freeways are at (or exceeding) their capacity. As speeds slow down, the space between cars shrink so there are actually more cars per hour at that pass a given point.
At 30-ish, the bumper-to-bumper distance is about 1/2 to 1 car length.
At 60 and above, the bumper-to-bumper distance increases to 2-3 car lengths.
Any slower, or faster (!) actually reduces the number of cars per hour past that given point.
Yeah, maybe you will remember those mind-numbing Driver's Education films we were forced to watch when we were 15 telling us that we should leave about 1 second per 10 mph between yourself and the car in front of you. Three car lengths in front of you at 60 mph! Like that's gonna happen on an LA freeway.

But this is because the city highways are absolutely packed with cars. Over 100,000 cars a day.

If you move to the countryside, then the concept no longer applies because the roadway is not at peak capacity. Lets say there is 1 car per day on this lonely stretch of highway. It won't matter if the speed limit is 30 or 150. 60 cars per day? 120 cars per day? 120 cars per hour? This is still "only" two cars every minute.

So, speeds "could" increase to a maximum speed of <nnn> (say, 180) mph on the open road, but such juicy spots will be few and far between. And even in the super-safe Model S P500D++, crashes are going to be spectacular and gory. We speeders still wouldn't care if it were only our own lives we risk, but there are others that will pay the price of our recklessness.

So I can see a non-automated maximum speed maxing out at 85-95, as this is a speed where human reflexes might actually still be quick enough to save lives. Mostly. A dedicated autopilot-enhanced lane (previously the carpool lane) could go faster, but why bother? By that time, we will have our own jet packs, flying cars, Hyperloop tubes, ...

-- Ardie

 

[brianman]

Just a random data point...

Looking and caring about speed limit signs (and law enforcement) is one of the most distracting things for me while driving. Model S helps in this regard somewhat, but if there were no speed limit signs or officers then I'd probably be a safer driver.

I'll leave it to the gallery to decide how to make both of those distractions go away entirely so that I can be a safer driver, and others like me.

 

[bxr140]

To offset the range penalty, the cars will follow each other very closely under machine control, so that they brake and speed up together.

I don't see that as the peak though.

Agree completely we'll see trains of cars, rotating the leader for efficiency like geese or bicyclists.

But...I think somewhere around 80-100mph will be peak speed:

1. On a relatively busy road (I-5 between socal and norcal, for instance), automated trains of cars won't be hard to come by, but in more remote places one can't rely on a number of cars to line up for efficiency. This limiting aspect is bounded by the future of capacity and charging speeds/locations.

2. Reaction time is still a safety issue for the random unexpected things encountered in the uncontrolled space of a freeway, like debris, blowouts, animals, etc. An automated train of cars on a freeway is very different than an aircraft flying in controlled space, a passenger train rolling down controlled tracks, or a hyperloop traveling in a sealed tube. This limiting aspect is very much bounded by physics, even with an automated system.

3. Automation is going to change the perception of time for a traveler and in fact the vehicle itself. Offsetting driving hours with things like work, sleep, family time, teevee, etc. will really drive down the priority for a driver to get to point B ASAP. Just as many Tesla owners find the supercharging experience refreshing and overall net positive even if they show up later than if they were in an ICE, full automation will have the same effect, probably tenfold. Travel time will find a priority somewhere behind time value and comfort. Then what happens when your vehicle is more like a long haul first class suite? How many people would choose to redeye themselves between the bay area and LA (for instance) over the other alternatives? How many would care if the ~7 hour Tesla trip (in today's time) took even 8-9 hours? How many would even prefer that? Imagine the efficiency of a redeye car-train traveling at 60mph...

This aspect really knows no bounds.

 

[david_42]

I suspect as more cars go electric, speeding will decline. Going from 70 mph to 90 mph doubles your energy consumption PER MILE. Or flipping it, you've cut your range in half. It would probably impact the total life-cycles of the battery as well. After paying another $10K to go from a 250 mile battery to a 300, you whack it back down to 150.

Not saying I didn't cross Montana with my foot to the floor several decades ago.

 

[Ludus]

Autonomy works a lot bettter if all the vehicles on the road are autonomous and networked. A few human drivers would prevent major improvements in traffic efficiency. I don't think it will be a decade after the first L5 vehicle fleets that a city bans human drivers entirely. Eventually driving will only be allowed on tracks or designated rural areas.

In that environment traffic will move very fast but there will be zero speeding. Traffic will be very lawful. Other crimes will be a lot harder too. No police chases. No getaway cars. Every street crime is seen by a dozen cameras in vehicles.

At intersections in an all autonomous city there won't be traffic lights, traffic will just flow through at full speed in all directions at the same time.

 

[theboom1]

Autonomy works a lot bettter if all the vehicles on the road are autonomous and networked. A few human drivers would prevent major improvements in traffic efficiency. I don't think it will be a decade after the first L5 vehicle fleets that a city bans human drivers entirely. Eventually driving will only be allowed on tracks or designated rural areas.

In that environment traffic will move very fast but there will be zero speeding. Traffic will be very lawful. Other crimes will be a lot harder too. No police chases. No getaway cars. Every street crime is seen by a dozen cameras in vehicles.

At intersections in an all autonomous city there won't be traffic lights, traffic will just flow through at full speed in all directions at the same time.

Don't make me laugh so hard. You think they are going to ever in anyone currently alive's lifetime going to ban humans driving? Lol not with all the people that would be opposed to it. They may make it so you cant drive in a car that doesn't have a high degree a automated safety system like auto braking and a much more aggressive lane keeping system that makes it so you can't drift off the road or prevents a lane change when it could cause a wreck. In fact if they mandate the auto braking and aggressive lane keep, they could probably raise the speed limits and on the interstate and make it like the autobahn with no speed limits (yay luda mode) if the systems were good enough. I mean if you can't run off the road, you can't make a unsafe lane change, the car auto brakes when you don't even at things you don't see, what else is there left to do? Presumable the majority of people still driving actually still want to so they will actually be paying attention and the people that don't will just switch on autopilot. Maybe the car car just go into autopilot when it senses you are not paying attention. They have some pretty fancy eye detection stuff now. No need to ban human drivers. If fact that's what they will have to do to accommodate the people that still want to drive. No traffic lights? Will not happen for so many reasons.