Slashdot Mirror
The Humans Crashing Into Driverless Cars are Exposing a Key Flaw (bloomberg.com)
schwit1 sends in a story from Bloomberg pointing out that the rigid adherence to traffic laws and overcautious programming have caused self-driving cars to rack up a crash rate twice that of an average human driver. "This may sound like the right way to program a robot to drive a car, but good luck trying to merge onto a chaotic, jam-packed highway with traffic flying along well above the speed limit. It tends not to work out well. As the accidents have piled up — all minor scrape-ups for now — the arguments among programmers at places like Google and Carnegie Mellon University are heating up: Should they teach the cars how to commit infractions from time to time to stay out of trouble?" While the autonomous vehicles aren't at fault in these crashes, their relative unpredictability on the road are nonetheless leading to more accidents than expected.
I have always thought that for automated vehicles to be a reality, ALL traffic has to be automated.
It takes almost an A.I. to be able to adjust to the random nature of human driving.
- Don't do what I do, it's probably not healthy nor safe. -
This should go both ways. People will need to adapt to the way automated vehicles drive (this would be helped by labeling them so they are easy to spot). Then automated vehicles should be given a set of exception to the rules and this would need to be legal, so the can override the regulations when the regulations are likely to create trouble.
When his defense asked, "Which computer has Jon Johansen trespassed upon?" the answer was: "His own."
People expect Caddys to drive slow and do weird things, because Uncle Harry is driving. Same for Priuses, because it's either Aunt Marge or some granola-head hippy doing his "hyper-mileing" thing. Problem solved :-)
Either that or put a sticker on the back: "This car rigorously obeys all traffic laws"
It might be safer in some situations, but you won't be able to buy an autonomous car that's programmed to break the law. The DoT will just never allow it, and I can't blame them.
Perhaps the robots and their rigid adherence is just making more obvious the flaws in many traffic laws that most people have known about for a long time now.
...and the middle lane.
Prove anything by multiplying Huge Number times Tiny Number
I'm sure there will be AI defenders who will question the assertion about a crash rate "double" that of average humans. But it doesn't matter. The point is that human drivers are idiots and drive in all sorts of unpredictable ways. They also tend to hate other drivers who operate in demonstrably safer ways (e.g., allowing plenty of space in front of them, not accelerating wildly just to stop 100 feet ahead in stop-and-go traffic, not zooming past a slower lane in a merge situation, but instead attempting a "zipper merge" at the same speed as the slower lane, etc). Of course, a lot of the less safe human behaviors also tend to be the reason for traffic snarls in the first place, but you'll have a hard time convincing most drivers of that, since they want to drive as if they are on a racetrack and somehow think that weaving back and forth to get into that tiny gap you've left in front for safety is going to allow them to get home so much faster (even if it's only 2 seconds earlier).
I imagine the biggest problem with having AI cars obey traffic laws strictly is not the accidents -- rather that it's going to lead to human road rage, which often leads humans to be even more irrational and drive in even less safe ways. Thus, while AI cars are still a minority on the roads, I'm not sure it will lead to a net improvement in accident statistics -- just as a "slow driver" on a highway can block up traffic, cause other drivers to drive unsafely around them, and ultimately lead to the potential for more accidents, even if that slow driver thinks they are being "safe" by driving the speed limit or a little below.
There was a study a few years ago about traffic in cities. They found that if all the drivers kept to rules that most cities would halt into complete grid lock.
People need to break rules to clear junctions, to pass cars that are stuck, and even force priority to not starve lanes going into a junction.
I travel by bus to and from work in Amsterdam, it is quite a long trip which includes traffic jams in the inner city. The bus driver needs to often break the rules to be able to pass cars, and force priority on junction because they are often stuck. Cars are backing up, cars are trying to make room.
WCPGW? There's a part of me that thinks, "We point lasers at planes!"
Happiness in intelligent people is the rarest thing I know.
Ernest Hemingway
Or the dummies who are speeding could just slow down and go the speed limit.
Get humans out of driving. Humans tend to break rules and that disqualifies them from being good drivers.
This is nothing more than a sophisticated from of "everybody else is doing it" argument that you get from small children. If the rules aren't working, the solution is to either enforce the rules better or to change the rules. Having everybody ignore the rules and not change them is the worst possible outcome. It creates a situation where things simply can't get better. Nobody can know the real effect of properly enforced rules so there's no data that can be used for improvement of the rules. What we need is better enforcement for human drivers. It's almost inexcusable that neither cars (nor trains) have automatic speed control systems that prevent exceeding the limit. Invariably somebody will point out the fantastical corner case where accelerating and swerving makes sense but those can be easily solved.
Not only is it not an infraction to drive in such a way as to save lives and prevent accidents, when you can save a life or prevent an accident, but it requires you to go against the suggested speed, or swerve into the left lane (even when the divider is solid) you are actually required to do so. That is the entire point of cars having a maximum speed of several times the maximum suggested speed is because you are supposed to speed in many situations to save lives.
Troll is not a replacement for I disagree.
So...everyone? The issue here is that people tend to drive as fast or as slow as the road allows, normally it's the common law speed limit. Humans can usually adjust to this, robots with strict rules can't.
It's difficult to say without having the telemetry from every collision.
If the one incident I've seen the data from is the typical case, then hell no. The driver rear-ended a Google car stopped at a stoplight without even slowing down.
On the other hand, if we're talking about merging, then that's possibly no. I don't drive big truck anymore, but let's face it: four wheel drivers just can't figure it out.
Merging is a bit of a difficult case because you can't merge if you haven't matched the speed of traffic. (Well, you can try....) So if the Google AI is trying to merge say onto 80 in downtown Chicago and is attempting to do it at 45 mph, which is the posted speed limit for pretty much all freeways in downtown Chicago, it will fail. I'd say this is the one case where the AI should flat out ignore the speed limit. Perhaps add some kind of heuristic to determine when it's a good idea to ignore posted speed limits.
On the other hand, if the on-ramps where they're testing are anything like the on-ramps in LA, well, the merging strategy I learned was get up to speed on the ramp, make a split second decision because that's all the merge lane you get, and pray to $deity. (Maybe they need to figure out how to get an AI to pray to $deity!)
Then the next step: just wait until bears start harassing these things. Every other car's going 80 in a 45, so the bears start regularly pulling over the Google car or production model autonomous car doing 75 in the 45. That will be an interesting legal battle since big city traffic never drives at the posted speed limit.
Which is a great theory, but the reality is that if the speed limit is set very low on a road for no apparent reason then a lot of drivers won't respect it, and unless you can and will enforce that limit strongly and consistently, that is unlikely to change. Putting the remaining drivers -- those who do want to be responsible and safe -- in a position where they have to choose between breaking the law and driving as safely as possible, is bad law-making.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
From TFA: "They’re usually hit from behind in slow-speed crashes".
If this is in fact the dominant accident mode, I would suggest that this is not such a big deal and will, over the long term, be self-correcting as the insurance rates for idiot non-automated drivers shoot up because they can't get it through their thick skulls not to tailgate other vehicles.
Indeed. If more accidents are prevented or at least the seriousness of them reduced from other aspects of enforcing the speed limit, then it's worth the occasional fantastical corner case crash.
Aside from not breaking the speed limit when necessary, the other source of rear end accidents is from a cautious response to the unpredictability of people in front. From the article: "But autonomous models still surprise human drivers with their quick reflexes, coming to an abrupt halt, for example, when they sense a pedestrian near the edge of a sidewalk who might step into traffic.". In my own driving, I take a cautious response to the unpredictabilty of people ahead and have been rear ended for it also. Until all cars on the road are driverless, probably the best solution would be for driverless cars to be designed with the assumption that they will be rear ended a lot, They should all have foam rear bumpers at least two feet thick to minimize damage when the inevitable occurs.
Which is a great theory, but the reality is that if the speed limit is set very low on a road for no apparent reason
Oh there is a reason, it just has nothing to do with safety.
I see it more as the cars failing to respond appropriately to surrounding conditions. The driving style of surrounding drivers is an essential condition that should be factored for.
When you meet a driver who will risk a collision rather than tollerate traffic infractions in others, what do you tell him? That he should stop driving by the rules? Hardly. No matter what you think about it, he's in the right and safe (from prosecution) as long as it's not him violating the rules. The same goes for driverless cars. Once you teach them to break the rules and one of them will get into an accident while being in violation, all driverless cars will be in trouble. Another positive result of this bevaior is that the more of such hardasses there will be on the roads, the safer the other drivers will be forced to drive in order not to cause a collision. In the end everybody will drive safer and less people will get hurt.
So...everyone? The issue here is that people tend to drive as fast or as slow as the road allows, normally it's the common law speed limit. Humans can usually adjust to this, robots with strict rules can't.
That is incorrect. People tend to drive as fast or as slow as enforcement of the speed limit allows. If authorities start enforcing the speed limit, the speed driven will decrease. Since there is no real penalty to speeding, people speed.
Having everybody ignore the rules and not change them is the worst possible outcome.
And yet, it's what we have.
Driving is a complex mix of the statutory, the habitual and the negotiated. Getting a car to obey the first is the easy part. Good luck with the other two. Especially the third. When a driver flashes his lights at you at a junction, what do you do?
The real world is messy and the corner cases kill you. If you're fortunate, you work in an industry where that's metaphor. If you're unlucky you work in an industry where that's a literal truth.
I had a dream, bright and carefree, but now there's doubt and gravity
Once we apply control system concepts to this, it becomes immediately obvious that any system of rigid adherence to traffic laws is a lot less efficient than more flexible system based on human-to-human interactions and learning.
Congratulations. This is the most amazingly elaborate rationalization for road rage I have ever seen.
Which is a great theory, but the reality is that if the speed limit is set very low on a road for no apparent reason then a lot of drivers won't respect it, and unless you can and will enforce that limit strongly and consistently, that is unlikely to change. Putting the remaining drivers -- those who do want to be responsible and safe -- in a position where they have to choose between breaking the law and driving as safely as possible, is bad law-making.
Actually, you don't need to enforce it consistently. You get as much compliance, but at a lower cost, if you haphazardly enforce it. If the driver doesn't know when it will be enforced, they will comply. It only takes the possibility of being caught that triggers the behavior.
"common law speed limit"
There is no "common law speed limit" in the USA. This isn't Britain.
If you think there is a "common law speed limit", you are self-deceived. (As well as a lawless and therefore despicable person.)
Yes, but so what?
If you want to know how these laws can be followed to the letter but robots still commit mass murder, read the Asimov short stories where those laws are taken from.
bickerdyke
which of cause the characters in the stories just never thought off because the assumed that those three rules were perfect.
bickerdyke
Can you just imagine being the first person on the road with a fancy new car that won't break the speed limit? Having people cruising by you on the freeway with a 15-20 mph closure rate? And then, when everyone is stuck at 55mph and can go no faster, it will be a glorious time. Rush hour traffic at all times of day, no ability to maneuver into another lane without slowing down. By all means, bring on our benevolent traffic dictators because it's about time we proved once and for all that yes, it is only speeding that causes accidents.
Personally I can't wait for an automatic car for my commute. However, when the last STi is built with a manual transmission and no nannies, I'm buying it and keeping it until the day I die.
Don't make self-driving cars shitty drivers just because everyone else is one.
1. Write all the code in Spark or at least Ada or another language with a similar safety record (e.g. Haskell, perhaps Rust). It can be a formally sound and static subset of C++, if it must be, but not just any C or C++.
2. Use only deterministic code, no dynamic memory allocation, fluffy A.I. heuristics or machine learning. I don't care how hard it is. I want real engineering based on real physics with provable security margins, guaranteed response times.
3. Formally validate the code in a theorem prover, as it is done for planes.
4. Let the whole design be reviewed and checked by external expert commissions, according to previously defined safety standards.
After that, we can discuss whether they are allowed to break the rules.
Why all that? Well, nobody should give a damn about safety statistics compiled by the companies who produce the self-driving cars, and it doesn't even matter whether they are safer than human drivers or not. What matters is that these companies have tons of lawyers, whereas the end-consumer has not. So good luck if you have an accident and it could be their fault. There needs to be some tight regulation like in the aviation business.
Once we apply control system concepts to this, it becomes immediately obvious that any system of rigid adherence to traffic laws is a lot less efficient than more flexible system based on human-to-human interactions and learning.
[Citation needed]
Your primary argument is that the benefit comes from "dynamic" systems that "are capable of adjusting to a wide range of situations." Except with more predictable adherence to traffic laws, you'd have a smaller range of situations occurring.
Also, in terms of "efficiency," it's pretty well demonstrated in traffic theory that traffic has "transition thresholds" kind of like fluid dynamic transitions between laminar and turbulent flow. When you have conditions like people trying to drive too fast and tailgating in high-density traffic, there's a greater chance of someone needing to brake suddenly, which often causes a chain reaction, and can generate a "traffic wave" that snarls traffic for 15 minutes in stop-and-go waves for miles behind. Similarly, in a merge scenario, often an efficient "zipper merge" might be able to happen at speed X if everyone were driving at a constant rate and merging at the appropriate moment. But if everyone instead is trying to drive at speed X+10 and tailgating to "not let the jerks in," it actually creates a bottleneck which results in the effective flow-rate through the merge to be at speed X-10, i.e., lower (and less efficient) than if everyone drove slowly and allowed adequate merging room.
In sum, there are a lot of common traffic situations where irrational human behavior is the primary cause of traffic snarls in high-density traffic. There are traffic flow studies that show even a small quantity (like 10-20%) of drivers driving more rationally (leaving gaps in front, avoiding crazing acceleration and braking, etc.) can actually "seed" a traffic snarl and cause it to break up and increase throughput. That, for example, is why have a large number of trucks on a highway during a traffic jam can often cause it to clear more quickly -- the trucks tend to avoid a lot of stop-and-go, and they need to allow adequate braking distance, so they effectively can help to clear a jam.
Anyhow, I don't buy the idea that the dynamic irrationality of human systems are actually more efficient, since the vast majority of traffic problems where you end up sitting in stop-and-go traffic for 30 minutes are caused by human error and accidents, usually introduced by violation of various traffic laws.
If the posted limit is less than the 85th percentile speed as measured by a survey in the last 5 years, then in California, it is legally a speed trap, unless several exceptions apply:
a) School zones are always 25 mi/hr
b) on freeways, the maximum speed is 55, 65, or 70, depending on the circumstances. Partly this is a holdover from the federal energy conservation laws of the 70s.
Since (b) is widely viewed as an artifact of older days, in practice, the posted limit is not enforced, rather it's a "what is safe". If you happen to get stopped going, say, 70, on a 55 stretch of road, the ticket will be for violating 22348(a).. the posted limit, rather than the safe speed. This is for convenience - no arguing in court about whether it was safe.
how do you enforce something 100% without going for a big brother solution? there are only so many cops.
Originally, the Interstate speeds were set by surveying the speeds of the cars. (I read that Engineers did this survey and on roads without speed limit signs--that us, unregulated. Other sources indicate that this is a common practice for setting speed limits on "new" roads.)
85% of the cars drive at the same (range) and that becomes the speed limit. The original Interstate speed in my state was 70. Kentucky was 75.
During the Carter-oil-embargo, the speed was reduced to 55 "to conserve oil". The USA Congress set the speed limit by law and USA "grants" were tied to the individual states reducing the speed limit. Years (decades) after the end of the Carter-oil-embargo, the limit was raised to 65 and a long while later to the current 70.
So, what was a realistic manner of setting speed limits has become a political football. Thank you Federal, State and Local Governments.
The is one town in Alabama that lowers the Interstate speed to 35 in that town's jurisdiction. Rather bizarre.
I love how strictly obeying traffic laws is called "unpredictability".
If an autonomous car can completely avoid accidents by taking corrective action that keeps it's behaviour within the law, then it should do.
But there will always be occasions that occur out of the ordinary. Take the obvious example of someone stepping out into the road - if you do nothing, then you are certainly going to crash into the person (and likely kill them).
Slamming on the brakes might cause a car behind to run into you, and it may not even be possible to stop in time.
Swerving may be the only option to avoid the person, but that takes you into oncoming traffic. You might hit a car head on. Or catch part of one, and spin and still hit the person in the road.
There are lots of permutations and possible outcomes, and staying entirely within the law may not always give the best outcome.
Predominantly, autonomous cars should keep to being within the law as much as possible, and if they do have to take some form of evasive action that stretches the law, they should be looking to get back to being within the law as quickly as the safely can.
But it doesn't make sense to completely shut off the option of going outside the normal limits, if the sensors are good enough and it is deemed to be the only way to avoid a collision.
It is also the same problem with automated law enforcement such as speed traps and red-light cameras.
Normally, cops are supposed to use some judgment before handing tickets, machines don't. A typical example is that you are supposed to make way for emergency vehicles, which sometimes involves breaking the usual rules. For example, if you cross a red light just enough to make way for an ambulance, a human cop won't ticket you as you did what you had to. A camera doesn't care. And it starts becoming a problem as less and less people make way for emergency vehicles if it may involve getting auto-ticketed.
If you think keeping up with the flow of traffic and/or occasionally driving in excess of the posted speed limit makes one a despicable person, I hate to imagine how you cope with everyday life surrounded by people you hate.
their relative unpredictability on the road are nonetheless leading to more accidents than expected.
Isn't the opposite way?, aren't self-driving stricly predictable?, isn't fault of those who don't follow the rules?. They should keep being like that, and people who doesn't follow the laws should be banned from driving, simple as that.
Believe it or not, "everybody does it", or, called by the official term, "customary law" is part of the anglo-saxon law system.
Drafting new rules is non trivial I guess, as they have to allow cars to do what the humans do, as well as still being understandable by humans, so that the humans stay legal, and the humans know what to expect from cars.
The problem is far more complicated that you realize.
It has already been shown that if everyone followed existing laws perfectly, traffic would grind to a complete stop. So obviously you need to change the rules, right? But, trying to change the rules to accommodate every possible situation will simply result in a mess that's even worse than what it is right now, because those "fantastical corner cases" are much more common than you think.
They're not going to though are they?
People can't be controlled in this way unless we fanatically enforce the speed limits.
"their relative unpredictability on the road are nonetheless leading to more accidents than expected."
So, self-driving cars don't use turn signals either?
It has already been shown that if everyone followed existing laws perfectly, traffic would grind to a complete stop.
Oh, come on. Who showed this? Where?
Which is a great theory, but the reality is that if the speed limit is set very low on a road for no apparent reason
Oh there is a reason, it just has nothing to do with safety.
Or there is a reason, it has to do with safety, or with optimising throughput, or some other valid concern, but that reason is not obvious to every dummy driver on the street. I don't trust every guy who owns a pair of pliers and a power drill to have a go at my dental care. So why would I assume that I know better than the experts which speed limits are optimal for a given set of goals?
Stephan
Oh there is a reason, it just has nothing to do with safety
Surely this article shows that the speed limit is indeed about safety. If you think that you actually driving at a safe speed and then run into a car that is travelling at the legal limit then obviously you were driving too fast to be able to avoid a hazard on the road. If you travelled slower then you would have more of a chance to see the car, correctly gauge it's speed, and then stop before you ran into it.
"...their relative unpredictability on the road are nonetheless leading to more accidents than expected."
Quite the opposite - they are predictable in that they follow all traffic laws - that's very predicatable. It's the unpredictability of human nature that causes the accidents. Should these robots be taught to break the law in order to conform with the behavior of their more chaotic human counterparts???
The article brought up another ethical question - the just one posed in Issac Asimov's book, "iRobot" - where the robots calculated the probability of survival and made the cold, calculating decision on who to save. In the case of a self-driving car and self-driving bus full of children, should the both vehicles decide to save the children or, simply, their occupants. Clearly, the latter would result in serious consequences if they both take the latter approach.
So more cops to that the robot cars are safer?
That's a head scratching response...
There are two types of people in the world: Those who crave closure
I am not sure why you brought road rage into this discussion, unless you consider routine traffic violations an example of road rage.
To make my argument easier for you to understand: Strictly adhering to a speed limit doesn't make you a safe driver. Driving according to road conditions, your alertness level, and your car's ability does. For example, by backing up traffic by strictly following 60 mph speed limit on a sunny clear day on a straight road you probably killed 0.001 humans. If you do this regularly, then over years you can rake up higher death toll than a drunk driver.
Typical incident, I want to get out the bus at the same time someone else also tries to get out. "Oh! I am sorry, please go ahead!", "no, its my fault, *I* am sorry, you go ahead", "after you.." , "OK thanks...". We walk to the park-and-ride lot, drive out and at some loop inside the parking lot some car cuts ahead of me, horns blare and the driver is flipping the bird at me. Wait, it is that same polite gentleman who insisted on waiting his turn to swipe the bus pass at the RFID device.
What happened? The only rational explanation is that the polite guy has been transported to some other universe. And some heartless cruel changeling scrap dealer from Jakku who exploits scavengers has come in and transmorgified into that polite gentleman's form.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
Yep, the headline is right. Humans crashing into cars that are obeying all the rules to the letter does expose a flaw: in the rules. They are impractical to the point of being almost impossible to follow safely. I guarantee if you got stuck behind someone on the road following all the rules religiously you'd be pissed off until you could get around them. That's not a problem of them, it's a problem of the rules no longer making sense (if they ever did).
This is nothing more than a sophisticated from of "everybody else is doing it" argument that you get from small children. If the rules aren't working, the solution is to either enforce the rules better or to change the rules. Having everybody ignore the rules and not change them is the worst possible outcome. It creates a situation where things simply can't get better. Nobody can know the real effect of properly enforced rules so there's no data that can be used for improvement of the rules. What we need is better enforcement for human drivers. It's almost inexcusable that neither cars (nor trains) have automatic speed control systems that prevent exceeding the limit. Invariably somebody will point out the fantastical corner case where accelerating and swerving makes sense but those can be easily solved.
You sir, are part of the problem and not the solution. For one thing, it is perfectly reasonable and acceptable to exceed the speed limit in order to safely merge into traffic. If you end up directly next to a car and need to merge then you have two options. One is to speed up and one is to slow down. If you're already going the speed limit then the safest option is not to slow down. You can see in front of you and next to you much more clearly than behind you. So why would you stick to a strict interpretation of the speed limit in order to merge? It's more dangerous than speeding up a few miles per hour, pulling into the gap you can see, and then driving the speed limit. Your inflexibility on the road is unsafe for yourself and everyone around you.
In fact, your strict enforcement of the rules is very difficult in the state of California. The state has a 'basic speed law' for any road with a speed limit under 55MPH. If you are not exceeding 55MPH then you may drive any speed that is safe for the road conditions (certain exceptions apply). So an autonomous car that strictly follows the speed limit could very much be a problem in situations where the basic speed law applies. Roadways are very fluid and dynamic environments. You have to have some leeway. Sometimes you need to be a little more aggressive. Sometimes you need to be a little more cautious. So yes, these autonomous cars should be able to temporarily ignore certain rules in order to increase safety. Of course, these autonomous vehicles can see behind them much more clearly than a human so the same safety guidelines may not apply to them as apply to humans.
In the UK you can be pulled-over if you are driving in the middle lane and there is no one in the outer lane (left-lane in UK, right-lane in US). In France there are special outer-lanes, which are clearly marked for people intending to drive for a single exit, so here the 'middle' lane is permitted for longer distance drives. I am wondering whether we could get a hybrid solution?
Road awareness campaigns would work too.
BTW the rules for these cars should be keep to the speed limit, but observe the traffic flow and adapt. Anything else is considered dangerous. The law enforcers hopefully are smart enough to realise this is the safe thing to do, especially when safety distances are being respected.
Jumpstart the tartan drive.
I'll admit it: I drive like an old person. Any day I can piss off a millennial in his BMW, is a good day.
You are welcome on my lawn.
I am not sure why you brought road rage into this discussion, unless you consider routine traffic violations an example of road rage.
I think it might have been you citing "honking and middle fingers" as examples of a "more flexible system of human-to-human interaction".
Strictly adhering to a speed limit doesn't make you a safe driver. Driving according to road conditions, your alertness level, and your car's ability does.
All of which are routinely mis-estimated by people who think they're better drivers than everybody else.
I'm training a new driver, and we have the same problem. She understands the rules of the road, but I spend a lot of time explaining to her where other driver's will expect her to behave differently than she expects based on those rules.
haven't rtfa but that seems like the screamingly obviously assumption. it's well documented that reven, er, "safety" red light cameras dramatically increase rear endings. I live in atlanta & try to be good about stopping for lights but experience has taught me to do what nascar refers to as "mirror drive" when I approach a light in traffic & go ahead & run it if person behind me looks like they're going to. I figure if there's a cop they're gonna get the last one anyway which is never me.
Strictly adhering to a speed limit doesn't make you a safe driver. Driving according to road conditions, your alertness level, and your car's ability does.
All of which are routinely mis-estimated by people who think they're better drivers than everybody else.
No disagreement here. Perhaps we should try German approach of autobahns and very strict graduated licensing?
Given various technologies like radar detectors and the Waze app, many drivers believe that they do have information about when it is being enforce.
I'm a leaf on the wind. Watch how I soar.
Conversely I don't get how autonomous vehicles are 'relatively unpredictable' if they are entirely unable to deviate from the laws of traffic.
Sounds to me like it's the humans that are unpredictable.
-=This sig has nothing to do with my comment. Move along now=-
The solution to this "problem" is quite obvious.
Highways need to be -all- autonomous vehicles. No manual control at over 50 kph (35 mph).
Don't take life too seriously; it isn't permanent.
I read long ago about a race-car driver teaching urban driving. He taught students to 'go with the flow' rather than compete for a place -- he'd make eye contact with other drivers, slow slightly to make a bigger space for someone to change lanes, wave and point to indicate where he was going, and almost never either accelerate or brake suddenly -- tap the brakes before stepping on the pedal to flash the lights. And at the end of the class he challenged the students to drive cross-town in rush hour, and he did as well. He was sitting in his car at the finish line waiting when the first of them arrived. He said you can flow faster than you can compete in traffic.
I'd always found that idea worked and when I read his methods I started applying it very seriously.
I'm in California. I get the momentary double take from other drivers when I do something gracious like the above -- the look twice at my hand signal to see if I'm really waving "come on over" or pointing "I'm moving that way" rather than something dumb and offensive.
And then, invariably, there's a blink and a smile and they make the flow work.
Another example -- when I'm behind another car, and that car ahead of me is coming up behind a slow moving truck, I'll put on my turn signal to change lanes early. It signals the driver ahead of me that, duh, we BOTH are going to have to change lanes soon, you might as well do it rather than be stuck behind that truck as I come up on your left. That almost always gets the driver ahead to move left in plenty of time for us both to flow around the truck. Otherwise, the contentious/oblivious driver will run up behind the truck then dart left into the passing lane as I'm coming up on them, endangering everyone. People do take a clue.
A third example -- I hit my 4-way emergency flasher if I see trouble or congestion or stopped traffic ahead way early, before I start to slow down and signal which way I'm going if I need to change lanes. When I don't do that, there's usually a pack of fast drivers coming along who will just blaze into the congested area and scramble for place trying to get through.
With the slight early warning, the whole pack slows down and people sort themselves out to flow past the obstruction.
This all works because there are PEOPLE driving the other cars. And it really does work. I get places faster with less hassle and people I've never seen before smile and wave at me on the highway --- because the little kindnesses propagate.
Try to be nice to a Google car? Why bother, there's nobody home.
I don't disagree with what you are saying, but I think you failing to consider very simple math - that if at 60mph you can get X cars through at peak, then at 120mph you can get close to 2X cars at peak. Therefore, it follows that drivers that are not willing to speed are constantly making traffic worse, while drivers that speed only occasionally make traffic worse.
I imagine a lot of traffic would go away if we set all highways to 80mph without drastically increasing road kill ratio.
To summarise the summary: people are a problem.
There will be the drivers who love to drive and will *never* give up their manual car versus having an autonomous car. It will be like the gun-nuts "You can have my 1967 Camaro when you pry it from my cold dead hands"....
Then the airlines will complain when people start using autonomous cars for long trips, and they sleep through they whole thing, wake up at their destination, as opposed to going through long lines and TSA hassles at the airport.
And the second there's a single accident that kills someone, the Anti-AI nutjobs will be protesting in D.C. "Them there robots is out to KILL us!"
And the complaints from Taxi/Uber drivers "Dey Tuk Our Jebs!" And don't even get me started on Truck Drivers, who seem to make of 50% of the working population of the USA..
We are clearly moving in the direction of having self-driving cars, but it's going to be a very, very painful and long transition. We may not even see it happen completely in our lifetimes.
If telephones are outlawed, then only outlaws will have telephones.
Finally, a rational post. Wish I had mod points today.
If you had read the bolded part and it's implications you would have discovered that your reply was unnecessary.
What he/she is saying is that by changing the rules to encompass situations you describe, an autonomous vehicle would never have to break any rules. Having an autonomous vehicle which bend or break the rules in certain situations is a sure recipe for accidents and getting sued into oblivion.
In other words, the traffic laws needs to be updated to take into consideration autonomous vehicles. As with all other emerging tech we have laws and regulations that are lagging behind and in some cases they are totally obsolete.
--- Reality doesn't care about your opinions, it happens anyway and if you are in the way you'll get squished.
If you're not overtaking, you have no business being in that lane. That's what the outer lanes are *for*. It's people who don't understand that simple rule that cause accidents. Undertaking is illegal e.g. in the UK for a reason.
Okay, sure, let's pause development to put the three laws into the core. In order to have priority, they'll be down by, what, the assembly language libraries? Machine code?
So now you need to take the abstract concept (harm) and have it EXHAUSTIVELY defined. And of course, you can't define it using any of those undefined meatspace terms. It takes a while to translate "very fast" or "very hot" meaningfully in binary. Especially vague mental constructs like "safety" or "danger".
You'll be doing a few trillion lines without high or mid level languages. Be sure to consult philosophers so the concept transcriptions are airtight, otherwise you get "protect humans from themselves" and people thinking the laws failed (they work perfectly - as coded).
This post also goes for you other normals who think a range sensor can comprehend lives-at-risk runaway trolley bullshit. It's going to react to the first unusual condition, the way it's coded to, and so on. Magic AI doesn't exist yet, it's all still pseudo AI following code and more code. CODE YOU CODING COWS.
I'm about as big of a proponent of a less intrusive government as you'll find, but there are times when I believe more is better. This is one of them. I floated the idea past a few state senators I know here in Nebraska, that drivers be required to take and pass a thorough state patrol-approved driver education class every 10 years. People change, vehicles change, laws change, etc. And yet the way it is now, any idiot can take some extremely simple tests at age 16, get handed a little card and suddenly they can operate a motor vehicle for the REST OF THEIR LIFE with no further training or education. Ever.
I cannot think of any other discipline in life where there is never, ever any additional education/training. But I was shot down by my state senators (from both sides of the ideological aisle) for what I considered superfluous reasons.
I am not sure why you brought road rage into this discussion, unless you consider routine traffic violations an example of road rage.
As the previous post said, I think honking and using middle fingers (some of your examples of efficient communication) are generally indications of aggressive driving behavior, commonly known as "road rage."
Strictly adhering to a speed limit doesn't make you a safe driver. Driving according to road conditions, your alertness level, and your car's ability does. For example, by backing up traffic by strictly following 60 mph speed limit on a sunny clear day on a straight road you probably killed 0.001 humans. If you do this regularly, then over years you can rake up higher death toll than a drunk driver.
I think you need a lesson in causality and culpability. Generally speaking, the driver who is going 60mph in a scenario where the law compels him to is not creating this "death toll." It is the people who drive more unsafely around him (tailgating, swerving to pass, etc.) because they aren't patient enough to drive safely within the posted limit. It's perfectly legal and in fact legally encouraged to drive at or below the speed limit, particularly if in the right lane (or one of the right lanes on a multilane highway). I do agree with you that it is unsafe for someone to drive the speed limit in some cases in the left lane, and there are generally laws prohibiting that. But if the slow driver is not in the left lane, and if common practice among most of the cars on that road is to drive faster and the authorities have not raised the speed limit, then the culpability may be jointly shared by those who posted the limit and the people who drove in crazy ways to be able to go faster.
I will freely admit that I frequently go over the speed limit, particularly when the general flow of traffic around me is going faster. But I have no illusions that if I end up in a serious crash by speeding that somehow a slower driver was magically the cause of it. When I do encounter a slower driver, I either drive a safe distance behind him or pass when it is safe and traffic is clear. That's what a "safe driver" actually does when encountering a car driving the speed limit -- and if you do otherwise and have an accident, it's YOUR fault.
If, as sometimes happens, I see a hoard of jerks driving bumper-to-bumper, tailgating, and weaving while trying to pass a slower car, I generally back WAY OFF behind them, because they are the ones causing the hazard, and I've actually witnessed accidents happen in those scenarios, so I want to be as far behind them as possible.
So more cops to that the robot cars are safer?
That's a head scratching response...
Ironically, yes. More cops enforcing the traffic laws will make non-autonomous vehicles safer to operate. Since these non-autonomous vehicles are what are causing the high accident rate with autonomous vehicles, then the autonomous vehicles will be safer to operate, too. It's simple if a then b; if b then c; therefore if a then c.
If it is not safe to obey traffic laws, there is either a problem with enforcement or with the law.
Funeral directing is illegal in the UK?
There are other credible reasons for keeping traffic speeds down apart from safety, but if the average driver can't see any of them applying, they are unlikely to respect the law.
I have long wondered about an alternative system of speed control for our roads, which I first saw proposed by one of the driving organisations: if there is a genuine reason to reduce the normal limit then there should be a highly visible sign introducing the new limit, stating the reason, and also showing a clear marker if it's being enforced by a camera. I also wonder whether drivers should be entitled to challenge any prosecution for breaking that lower limit on the basis that the stated reason didn't apply.
My suspicion is that a lot more drivers would respect lower limits if, for example, they said "School" and only applied within a reasonable distance of an actual school. I suspect a lot of drivers would also respect genuine reasons like "Workers near road" if the limits and warnings were only used reasonably close to where there actually were, as opposed to for example dropping the limit for a 5 mile stretch of motorway with no-one working anywhere as we see all too often today. And for non-safety issues, such as driving through a quiet area, which might not always be obvious to someone driving past, an explicit sign asking for considerate driving surely can't hurt.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
So when it is going say 10 over the speed limit to keep with traffic and gets into a crash due to some unforeseen circumstances that maybe a bit slower wouldn't have had an issue, who wants the liability that it was indeed breaking the law, and was programmed to do so intentionally?
Lawyers would have a field day before and after if they are let off the hook... so minor infractions don't matter as proven in the case of autonomous cars...
Please explain how you would build the infrastructure to communicate speed limits to cars. Please include who would pay to build it and maintain it, how cars would get updates, and how the system would work internationally.
Have you even been in a car this decade? The system that you describe is already in place in cars today. We use them to help us navigate, just like the robots do.
I'm typing this on my phone, and it has two applications that can provide comprehensive maps that show exactly where I am and which way I am facing. I can look up maps for cities around the world. I have even used Google Maps to navigate a bike path in the dead of the night. It's like rally driving with the phone telling me which way to turn next. Scary as hell, but fun!
Having speed limits on these maps is trivial. Having context sensitive speed indicators changing with traffic conditions if you use the GPS data from the other cars ahead of you.
You take a sledge hammer approach to law enforcement. More traffic cops don't make the road safe. People learn where the speed traps are and simply slow down in those areas. It is as Colin said, people tend to drive as fast as the road and their skill level allows for.
From experience in the Netherlands, that point seems to be at about 1 ticket/year per driver.
Over here the police write about 8 million speeding tickets/yr, and there are about that many cars registered in the country. Over the past few years, the number of people I've seen speeding has plummeted, and the number of people driving at (speed limit -10 to 20 km/h) is large enough that it's becoming impossible to drive at the speed limit because there's always a slowpoke in front of you.
Generally speaking, the driver who is going 60mph in a scenario where the law compels him to is not creating this "death toll."
I disagree. Even if we ignore slow driver's culpability in the increase of accidents caused by other drivers, there is still an issue of slowing down all other traffic. If you just slow down entire highway by 1mph by your driving, then that alone would by sufficient to attribute road carnage element I mentioned.
There have been a few studies showing that speed traps are dangerous. People slow down suddenly when they become aware of police in the area, increasing the rate of collisions. Excessive enforcement, oddly enough, leads to more loss of life.
Support my political activism on Patreon.
"autonomous vehicles aren't at fault" -- this needs to be revised. Doing something totally unexpected can be considered as doing wrong. I see a car going ahead of me and also I can notice there is sufficient distance in front of that car. If he(that car) suddenly brakes too hard, it will surprise me and I can very well rear-end him. Is it my fault? yeah law can say I should've been careful. But we humans expect something what is reasonable. There is no reason why the car in front could not have done a slow/gentle slow-down. Basically when you drive, you don't surprise your fellow drivers. You drive predictably. [Even with wild-life it's true, you can walk, say in african savanna.. a lion/snake is likely to attack you ..when they are surprised.. that's why you create noise..announce your presence..so they know you are around ..and also gives them time to conclude you are safe/not a threat to them]
There's a middle ground to be had. Sure there are a lot of ubiquitous bending of rules that should be addressed (amending or enforcing). Speeding is among those. There are an essentially infinite set of 'exceptional' circumstances that are relatively rare in occurrence, but critical when they occur. Having laws to cover that would require an infinite set of laws. One could say you could get close enough with more and more complex laws, but at some point the law becomes outright impossible to understand, well before the laws cover all the circumstances.
XML is like violence. If it doesn't solve the problem, use more.
I think it's awesome. I had a friend at church complaining that he got a ticket for 65 (the speed limit) in the fast lane. I gave him zero sympathy and told him to move over for faster traffic.
Peter predicted that you would "deliberately forget" creation 2000 years ago...
Driving is and always been the most intense social activity we will engage in on a daily basis. We cooperate and even conspire with each other to break the law in order to get to our destinations quickly and safely. I doubt they will be able to program the complex second by second social decisions we make every day. I think Google now thinks the same as they have separated out the autonomous car part of Google, probably because they expect it to be sued into oblivion eventually by the families of autonomous car killed human drivers. If an autonomous car breaks the law and someone dies, the record is there in the database. Present that to the jury and watch the money flow. If they want autonomous cars to be on the roads of this country with humans, a corrupted and incompetent Congress will make sure we have to accept a certain number of machine-caused deaths and maimings. The Congress will have to strip human legal rights to allow widespread use of killer machines on our highways. Google has the bribe money to bring this about.
Another point: OSHA doesn't allow a certain amount of machine-caused deaths and maimings in our industries, why would it here?
E Proelio Veritas.
In California we have clearly marked school speed limits and construction speed limits. These are the most ignored limits according to cops, so I believe your theory to be incorrect.
Peter predicted that you would "deliberately forget" creation 2000 years ago...
Mostly because experts are people who have little slips of paper telling us their opinions are more valid than the opinions of others. It's the inversion of the American Cargo Cult ideal: while some people believe their opinion is as valid as any other opinion and that people who are well-researched just have an agenda, others believe that anything written by experts is Holy Writ and infallible.
Socrates had an issue with this.
Socrates figured people should ask questions about *everything*. We have a world where people get a glance at something, make up some stuff, and then stop asking questions about it and presume anyone who disagrees with them is a moron; on the other end, someone gets a certification telling us they know about something, so we stop asking questions and assume anyone who disagrees with *them* is a moron. Socrates wanted us to question ourselves *and* the experts, to find out why the experts disagree with us, and then to decide if the reason for disagreement was valid, or if we could even assess it with the information we had. Do that enough and you'll start realizing the experts are wrong--they're always wrong, just less-wrong than all of their predecessors.
Support my political activism on Patreon.
Oh there is a reason, it just has nothing to do with safety.
Or there is a reason, it has to do with safety, or with optimising throughput, or some other valid concern, but that reason is not obvious to every dummy driver on the street.
The throughput argument bears repeating. Many drivers don't understand this, but sometimes you can get more people through a bottleneck (AND have fewer accidents) if everyone drives more slower at a constant speed than if everyone is trying to drive faster at the same traffic density. This is particularly true when you have a high variability of vehicle speeds, like in a mountainous area where trucks are forced to go slower or in an urban area where frequent incoming and outgoing traffic at exits often travel at different speeds from the rest of the highway.
For example, if you're driving on a highway through an urban area and they lower the rush hour speed limit to 45 mph (some areas now are adopting such dynamic speed limit signs), the idea is that if cars actually go 40-45 mph, the road will actually be able to handle the amount of traffic while also allowing all the people merging on, getting off at exits, changing lanes, etc. at a safe speed.
If, instead, everyone tries to drive 65-70 in the same area, what can happen is that the merging or changing lanes will eventually cause someone to cut someone else off, which causes sudden braking, which then causes some tailgaters behind them to brake suddenly, others follow and overcompensate because they were going too fast and suddenly see much slower cars, and within a few minutes you have a "traffic wave" of stop-and-go traffic backed up for a few miles which might take a half-hour to resolve, where throughput is dramatically reduced. (How many times have you gotten to the end after sitting through 10 minutes of such stop-and-go traffic waves, and there's nothing there -- no accident, no merge, etc.? This is often the kind of thing that happened.)
At a slower speed, the slower car may not have been forced to "cut someone off" in the first place, or if he does, the impact of a bit of braking may not cause such massive changes and overcompensation. Traffic thus recovers faster and throughput is maximized.
It sounds more like a flaw with traffic laws.
You're right. For example, the safest speed on the highway can be described as "the number of cars I passed equals the number of cars that passed me". This kind of logic should be integrated into the software.
Dear Microlimp: I give you 2 valid product keys for win7 and you reject both of them. Piss off you wankers!!!
It's almost inexcusable that neither cars (nor trains) have automatic speed control systems that prevent exceeding the limit.
Trains don't have automatic speed control systems???? *cough*
Please see the rather short english article on Train Protection Systems (or the much longer German version) to enlighten you. It might be the U.S. track system is no longer up to standards but that doesn't say anything about trains in general.
If you had read the bolded part and it's implications you would have discovered that your reply was unnecessary.
What he/she is saying is that by changing the rules to encompass situations you describe, an autonomous vehicle would never have to break any rules. Having an autonomous vehicle which bend or break the rules in certain situations is a sure recipe for accidents and getting sued into oblivion.
In other words, the traffic laws needs to be updated to take into consideration autonomous vehicles. As with all other emerging tech we have laws and regulations that are lagging behind and in some cases they are totally obsolete.
Did you read his entire post? He wants to not only make the rules stricter for computers, but also humans. He wants to use GPS location data or something like that to have a speed limit specific governor on speed. So no, he's not saying that we need to change the rules to allow flexibility. He's saying that we need to make the world much less flexible for everyone. That just does not match reality.
I'm tired of morons thinking they understand traffic engineering. The speed limit should be set at around the 80th percentile of car speed on the road. If the 80th percentile is 20mph faster than the speed limit, the speed limit is broken; hell if its 10mph slower than average its being set at about the 30th percentile likely; so 70 percent of people are speeding on that road. Time and again studies show that most people drive a safe speed for the road and conditions regardless of the speed limit, whether faster or slower.
You on the other hand think that the speed limit is set by some omniscient being that knows whats best for the road and sets it and then the majority of people, being sheep, break it and lead to evil consequences. Because hey OBEY the limit, its not a retarded limit that needs fixed, because YOUR BETTERS SET IT.
I'm sure you'd prefer your cars designed by some randomdood since you prefer your roads and speed limits not designed by a qualified engineer either.
No, what the article shows is that, if 2 cars are moving at very different speeds, it is less safe.
It should be equally safe for 2 cars to be both moving at 70 mph, or both at 30 mph.
The problem is that, when the prevailing traffic, is going 70mph, human drivers (mostly) recognize this, and speed up to merge into traffic. But the AI is forbidden from breaking the speed limt, so it tries to merge into 70mph traffic at 55mph, which is risky
This could be fixed equally well by making the human traffic slow down + stay within the speed limit. OR, by allowing the AI to do what humans do, and adjust to fit the speed of the prevailing traffic, regardless of the speed limits.
The problem is they're programmed to be dead right. The thing about being dead right in traffic; you're still dead.
Another way of looking at it is that predictability is an important factor. Almost regardless of speed, someone behaving unexpectedly can cause issues. Human drivers ironically tend to offer clues of their unpredictability, allowing other drivers to flag them for extra attention, while self driving cars probably does a lot less of that (stuff like weaving a bit, acting uncertain, blinking the wrong way etc.).
Of course, if all cars were self driving, the issue would vanish. I suppose it would also lessen as more and more self driving cars arrive on the roads, effectively training other drivers in how they behave.
I disagree with everything you are saying, and I'm convinced you should consider not advising people on simple math. Safe highway driving intervals are measured in time, not distance -- maybe you once came across the "two second rule"? Doubling the speed doesn't increase throughput by a factor of two: if people are driving with any concern for their lives it won't budge. Your other argument in your earlier post, about slower people causing accidents, exists only in your rationalizing world. (I have worked in motor vehicle accident investigation, and am a mechanical engineer.)
The subject who is truly loyal to the Chief Magistrate will neither advise nor submit to arbitrary measures (Junius)
"of cause", "thought off", "the assumed"
I'm not usually one to nitpick on spelling, but this was a bit much on such a short clause.
There's nothing like $HOME
Mostly because experts are people who have little slips of paper telling us their opinions are more valid than the opinions of others. It's the inversion of the American Cargo Cult ideal: while some people believe their opinion is as valid as any other opinion and that people who are well-researched just have an agenda, others believe that anything written by experts is Holy Writ and infallible.
Socrates had an issue with this.
Well, for one thing, there is a difference between expert and "expert". But then, typically even an "expert" has spend more time and thought on a problem than Joe Schmoe the average driver. From my cyclist perspective, Joe Schmoe does not spend much brainpower on anything ;-).
Socrates figured people should ask questions about *everything*. We have a world where people get a glance at something, make up some stuff, and then stop asking questions about it and presume anyone who disagrees with them is a moron; on the other end, someone gets a certification telling us they know about something, so we stop asking questions and assume anyone who disagrees with *them* is a moron. Socrates wanted us to question ourselves *and* the experts, to find out why the experts disagree with us, and then to decide if the reason for disagreement was valid, or if we could even assess it with the information we had. Do that enough and you'll start realizing the experts are wrong--they're always wrong, just less-wrong than all of their predecessors.
Socrates lived in a time when the sum total of human knowledge was a little fraction of what we know today, and this it was a lot easier to get up to the state of the art in several fields. Sure, we should question experts, but on average, the risk of rationalising our snap judgement with half-understood facts and simplified sound bites is not insignificant. The expert is not always right, but he usually comes closer to the truth than a non-expert.
Stephan
It's not silly. It actually lets you go faster, and more safely with less road congestion. Why? Because slower traffic ends up on the inner lane, going progressively faster to the outer lane. So it's actually not silly at all in heavy traffic. It means slow traffic doesn't impede faster traffic, and it means that the driver can make certain assumptions about the driving conditions, e.g. they won't be overtaken on the inside, which makes lane changes easier and safer. It makes driving on fast roads predicable since you know how the other traffic will behave.
Of course, there are some idiots who don't know how to drive and hog the middle lane, and are universally hated for it; they get pulled over by the police since it's actively dangerous to the other traffic--because it can force people to undertake.
When you look at footage of road accidents in countries without this rule, what's immediately clear is that the driving is far more erratic since it's an unconstrained free-for-all, and that makes accidents more likely, as well as reducing the effective safe speed on the road.
You just identified the real issue "following too closely behind the vehicle in front of you" perhaps if there was a legislative/punitive or technological solution to this, issues relating to "traffic waves" would be solved.
Sig withheld to protect the innocent.
We are nowhere near “A.I” yet. The “I” is intelligence. Intelligence is learning, changes in behavior, adaptability, altering conclusions based in previous
experiences. We know the brain reprograms itself as neurons develop new pathways and old ones wither over time as our skills and life experiences change. This adaptable dynamics is the essence of intelligence and implies that any truly artificial intelligence may – more likely will - outgrow or rewire itself out of any law, rule or human moray – including Asimov’s laws of robotics.
Till dynamic adaptable learning is possible, stop calling systems that follow logic branches ‘intelligent’.
The right approach is to capture 360 degree video of every other car, then sue the owners or insurance and every other car that makes contact for millions in repairs (these self-driving cars are expensive, right?) In other words, don't conform to a lawless society, force a lawless society to conform to the rules! It would also help if they'd flag all these cars with a huge florescent green flag on top with dollar signs all over it, to warn people to pay attention! Other approaches would be electronic countermeasures -- hack the cell phones of everybody in the vicinity and post messages telling them to stop texting and/or shut up and drive!
I've abandoned my search for truth; now I'm just looking for some useful delusions.
If they are in the rightmost lane, traffic should be travelling within 10mph of the speed limit. I assume these vehicles are programmed to stay out of the fast lane!
I've abandoned my search for truth; now I'm just looking for some useful delusions.
Artificial intelligence is no match for real human stupidity!
I've abandoned my search for truth; now I'm just looking for some useful delusions.
Need better speed limits also unofficial lanes need to be looked at. When it's open don't even think about trying to just do the limit on I-294 I-290 I-94 I-90 I-355 or even LSD.
Like the mini medians that people use as an center left turn lane.
Using the shoulder to pass a waiting left turner on 2 lane roads.
The unmarked parking / though lanes on some the road that are not as big as real 4 lane roads but are used as an 4 lane road.
Using the striped median as an extend left turn lane.
If an autonomous car gets a ticket, who is responsible for paying it? Whose driving record does the offense go on? Obviously the cars being tested actually have passengers, are they programmed to pull over for emergency vehicles and wait patiently to receive tickets, or does the onboard technician have to intervene?
I've abandoned my search for truth; now I'm just looking for some useful delusions.
The work zone 45's need to go or be only for places where it makes sense like local roads not inter states where there is a wall from traffic and the workers. Work zone 55-60 are ok but still at times where there are no workers.
School zones should be limited to local roads and not a road that is 45 MPH that drops to 20 when there is a near traffic light that the kids can use. There are lot's of school zones where no hours as posted and there is no school near by.
This is exactly why control of cars needs to be out of human hands.
Optimal flow of traffic is too hard and large a problem for any driver to determine what they should be doing from their point of view, let alone the expectation that each driver is capable and conscious enough to make 'non obvious' moves, like going slower.
I have no doubt that with some algorithms and real time traffic stats that we'd all get to where we wanted to go much faster and safer than we can now, but we have to take our human hands out of it.
I don't understand why the pride issue or whatever either, there is no way that over the 70 years or so we'll probably be on the roadways that we can guarantee full cognizance at every given moment of driving. We all get distracted, mad, happy, surprised, excited, and each mood will affect our driving.
Eh, I'll never see it in my lifetime, but I tell you it'll be a much better world of transportation when computers are getting us from point a to b.
Socrates lived in a time when the sum total of human knowledge was a little fraction of what we know today, and this it was a lot easier to get up to the state of the art in several fields. Sure, we should question experts, but on average, the risk of rationalising our snap judgement with half-understood facts and simplified sound bites is not insignificant. The expert is not always right, but he usually comes closer to the truth than a non-expert.
That's no excuse for turning off the thinking part of your brain and being a twit. You can accomplish greatthings when you do a little thinking. In my case, I have economic theories that don't match with mainstream politics or contemporary economics, such as the observation that labor costs affect employment rates--one of the reasons I push for a Citizen's Dividend instead of a minimum wage and public-aid system, which lowers taxes and has positive impacts on retirement funding. This is also why I see the growing gap of income inequality as a good thing, since I can make use of it to cut back taxes on labor, thus incurring all those nice benefits like lowering the minimum price of all products, increasing employment, and possibly causing the full-time work week to shrink to 32 hours (figure that one out).
I don't have a fancy economics degree; I've cataloged a ton of information in my head, and at some point started organizing it. This is what happened. There are holes; I don't get how exchange rates work, among other things. I've come up with a few rough theories for those, and need more time to work them into something that fits in consistently with existing observations. It'll never be wholly correct, but it's not right until it appears to work under scrutiny.
It gets checked against contemporary theories, looking for observations I haven't made, looking for holes, looking for places where they were wrong; often I realize these people weren't wrong, but rather painfully close, like with Adam Smith asserting that the only way to work more efficiently is to divide tasks into smaller fractions spread among more people, rather than to invent new tools and techniques that may optimize processes (assembly line vs cellular manufacture: same tools, same laborers, different floor plan, less wandering back and forth carrying half-finished parts) or simply figure a way to get an end result in fewer major steps (technically, *consolidating* labor. Smith wasn't wrong; he just veered off at the end, observing the improvement of efficiency by doing more work with the same labor and then specifying it as division, instead of just the raw observation that more production per labor time creates wealth. He also was writing a theory of value, trying to explain the correct price of products instead of describing economic movement in general.
One thing I've learned in my life: math is a tool, not a framework. If people try to define non-mathematical things by math, reject that and look for a plain-words explanation. It probably won't be a simple one--simple explanations just gloss over complexities--but the mathematics behind quantum physics aren't as important as the basic concepts of how quantum mechanics works. If you know all the math, but not the mechanics, you're a Chinese Room: someone can hand you a paper with a lot of squiggles, you can pull out your dictionary and write down a correct response, but you don't really understand the conversation you're having. If you have the mechanics but not the math, then you can't be an engineer; you *can* be a theorist. Theorists are the ones who figure out where w
Support my political activism on Patreon.
I can't speak for California, but here in the UK, most drivers do tend to slow down reasonably for obvious hazards like kids outside a school, actual workers at the roadside, actually restricted spaces due to parked cars, and the like. I might suggest that what limited resources we have available for speed enforcement should be directed at the few fools who don't and instead persist in driving at silly speeds even when it is so obviously dangerous, rather than on yet more speed traps that frequently pick up drivers who were violating the law but probably not actually doing anything dangerous or inconsiderate. But of course that does require some degree of actual human involvement in enforcement, whereas cameras can rack up fines without needing the same.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
I believe their unpredictability comes from their inability to judge other normal human actions. For example, as mentioned above and in a few other articles I've seen, when driving near sidewalks autonomous vehicles have a hard time determining when a person is merely being an impatient jerk who thinks hanging his toes over the edge of the sidewalk gives him a real head start when the crosswalk light changes or is actually about to step onto the road. In those cases people might slow down (or not) but autonomous vehicles have been known to suddenly stop, even though the light is green.
So yes, the cars reaction is caused by an unpredictable human but not one that's driving and to other humans that impatient pedestrian, while being a annoyance for standing too close to the edge, isn't anything out of the norm.
Of course that's just my opinion...... you could be wrong!
I'm tired of morons thinking they understand traffic engineering. The speed limit should be set at around the 80th percentile of car speed on the road. If the 80th percentile is 20mph faster than the speed limit, the speed limit is broken; hell if its 10mph slower than average its being set at about the 30th percentile likely; so 70 percent of people are speeding on that road. Time and again studies show that most people drive a safe speed for the road and conditions regardless of the speed limit, whether faster or slower.
I have never seen such studies and would be interested in a reference.
Certainly I know that when I travel on windy mountain roads I can very easily exceed the posted speed limit with no risk of sliding off the curve, and many people driving those roads do exceed the posted limits (I have done so many many times in the past myself). It was a few months ago while reading a slashdot discussion about self-driving cars that I realized for the first time that many of the posted speed limits were there not because of a need to prevent cars from sliding off tight corners but because of limits to visibility and stopping distances. The "sea to sky highway" heading north to Whistler from Vancouver BC is notorious for traffic fatalities, largely I think because of this effect.
People drive at what they THINK is a safe speed for the road and conditions, but often are unaware of what that safe speed actually is. Modern cars are quiet and handle very well at high speeds and contribute to the driver FEELING like they are in complete control. Stopping distance increases proportional to the square of the velocity - so increasing your speed a fairly small amount can increase stopping distance by a much larger amount than most people think. So going around the corner at 50mph might not be safe while travelling at 45mph could give ample space to avoid problems blocked from view by the curve.
No thanks.
Your statement might be true (I have no stats in support or defence) but I'd argue that's a low-quality solution. You end up with a system designed to punish some people some of the time in order to get compliance. It encourages a system of traps for citizens. It encourages police to set up radar traps instead of using more or better signage. I'm really not interested in that sort of law enforcement.
There are tonnes of reasons it's pscyhologically negative. Particularly, young men like to push the boundaries of the world, so we end up with a system that doesn't consistently enforce the boundaries, but levies large penalties. This is the sort of trick you might want to use in prison, but on the roads?
Persistent enforcement with moderate penalties is really far better.
And if you want to live by that rule there are lanes on every highway meant for you, just stay out of the passing lanes.
Admittedly, I speed everywhere but I have no issue with people driving the speed limit if they stick to the right most lane(s), even if I can't get into the left lane because I'm unable to get up to speed to merge safely while following the turtle in front of me. It's only people who insist on driving at or below the limit in the left most lane when there is ample space for them to the right that annoys me.
Of course that's just my opinion...... you could be wrong!
Try those silly laws when there is traffic. Having a lane that you can't use except to pass becomes ridiculous when all the cars are averaging 5 mph or so.
At least in the UK, the rules are reasonable here and in any case no police officer is realistically going to be upset because you kept up with slow-moving traffic that is clearly flowing in lanes, maybe unless you're being a jerk by changing lanes all the time or something like that. See rule 268 in the Highway Code, for example.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
I imagine one of the problems (self-driving cars being rear-ended) is because when a light turns yellow or something happens, the self-driving car can react faster and might even apply the brakes much faster than a human would, so the normal space a person leaves to another human-driven car isn't enough. At least a symbol on the back indicating it's a self-driving car and a warning that it can brake faster than a human might help.
"I have never let my schooling interfere with my education." - Mark Twain
So why would I assume that I know better than the experts which speed limits are optimal for a given set of goals?
I don't know why you would, but I actually do know the relevant mathematics, I actually have seen details of some of the simulations our local traffic engineers do, and since my IQ is more than 5 I actually can laugh at the hilariously unrealistic assumptions they sometimes use. So when my local councillors try tell me that something is being done for the overall good, and appeal to the authority of their traffic engineers to support their case, and then I ask to see the data to back up that assertion, I can and will challenge it if they're talking out of the wrong orifice.
If you think the standard XKCD is the reality of how traffic engineers operate and your entire city's junctions are meticulously planned and co-ordinated, then either we live in very different cities or you are hilariously wrong. The planners here reportedly don't get as far as factoring in the effects of an awkward multi-mile detour for many months of major works when deciding whether it's worth doing those major works to make a relatively minor change in the road layout with the goal of quite modest long-term improvements in the efficiency or safety of the system.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
...they upgrade the AIs to be more aggressive, and they start revving the engine next to you at lights? Actually, that'd be amusing as all get out.
What does language have to do with priorities? Things have priority because you *tell* them to have priority, not because they were written in a more "primitive" language. Writing in a lower-level language is almost entirely a performance hack, and modern compilers are good enough that it usually doesn't even buy you much.
I agree that magic AI doesn't exist, and unusual situations will inevitably cause problems (hell, they often cause problems even with human drivers). But if we can virtually eliminate 70% of common accidents, then even if we make the 5% of really unusual accidents 10x more common we'll still be reducing accidents by 20%. (I'm assuming the AI roughly breaks even on the remaining 25% of inherently difficult or merely uncommon accidents. All statistics pulled out of my aether).
Of course there's also some consideration to be done on relative severity of accidents, but since we're purely speculating here that isn't really something that ca be productively discussed. Let AI-caused fatalities per mile reach something approaching human-caused ones and then we will have some data to work with. Until then we'd just be letting perfection be the enemy of progress.
--- Most topics have many sides worth arguing, allow me to take one opposite you.
While the autonomous vehicles aren't at fault in these crashes, their relative unpredictability on the road are nonetheless leading to more accidents than expected.
The autonomous vehicles are *entirely* predictable as they follow the rules - you know, the ones we humans are suppose to know to get a driver's license. It's the humans that are unpredictable.
... rigid adherence to traffic laws and overcautious programming have caused self-driving cars to rack up a crash rate twice that of an average human driver.
Um, humans are causing the crashes, so the crash-rate is on them, not the self-driving cars. The crash-rate for self-driving cars is "bagel".
It must have been something you assimilated. . . .
The issue here is that people tend to drive as fast or as slow as the road allows,
That is incorrect. People tend to drive as fast or as slow as enforcement of the speed limit allows.
Don't try passing off your misinformed opinions as fact. There is plenty of research on the affects of lowering or raising speed limits, and universally it finds most people drive at speeds they are comfortable at regardless of speed limits. One particularly broad study done in Virginia in 1992 provides some very detailed findings, along with the statistical relevance of those findings. It is quite clear that speed limits are far too low in this country and our roads would be much safer if they were raised on average 5-10 mph. Some important findings are listed below.
Posted speed limits were set, on the average, at the 45th percentile speed or below the average speed of traffic, even though raising speed limits in the region of the 85th percentile speed has an extremely beneficial effect on drivers complying with the posted speed limits.
Accidents at the 58 experimental sites where speed limits were lowered increased by 5.4 percent. The level of confidence of this estimate is 44 percent. The 95 percent confidence limits for this estimate ranges from a reduction in accidents of 11 percent to an increase of 26 percent.
Accidents at the 41 experimental sites where speed limits were raised decreased by 6.7 percent. The level of confidence of this estimate in 59 percent. The 95 percent confidence limits for this estimate ranges from a reduction in accidents of 21 percent to an increase of 10 percent.
At sites where speed limits were raised [by 5, 10, 15, or 20 mi/h], there was an increase of less than 1.5 mi/h (2.4 km/h) for drivers traveling at and below the 75th percentile speed. When the posted limits were raised by 10 and 15 mi/h (16 and 24 km/h), there was a small decrease in the 99th percentile speed.
Lowering speed limits by 5, 10, 15, or 20 mi/h (8, 16, 24, or 26 km/h) at the study sites had a minor effect on vehicle speeds. Posting lower speed limits does not decrease motorist's speeds.
Raising speed limits by 5, 10, or 15 mi/h (8, 16, or 25 km/h) at the rural and urban sites had a minor effect on vehicle speeds. In other words, an increase in the posted speed limit did not create a corresponding increase in vehicle speeds.
-- All that is necessary for the triumph of evil is that good men do nothing. -- Edmund Burke
Actually, it does none of that. Police used to issue traffic violations and speeding tickets without having traps. They would just be cruising on their patrol route and if they saw an infraction, they pulled the person over. What has changed, however, is that police staffing has been reduced because of erosion of tax basis and there are now fewer officers per 1,000 citizens and those remaining officers are tasked with other things than traffic. In short, with reduced funding the police have to choose between protecting property or streets and property wins. Police traps, at least in the US, didn't become really popular until the 1980s when budget cuts were enacted under the guise of trickle down economics. This left many police (and fire) departments under funded. So the police turned to traps, not because they are more effective, but because they are cheaper.
If you want safe traffic, somebody has to enforce it because, one only has to look at the speeds people are driving and the lights they drive through to see that individuals don't follow the rules without a consequence. OTOH, if you are okay with the current state of things, then nothing needs to be done.
Heh, it's going to take multiple teams of well trained (some probably genius level) engineers to formally prove what every man in the land understands at 16: the traffic laws cannot be taken as strict instructions, because they do not function as written in many cases.
Anyway, I'm glad this hangup is happening now- if it happened after adoption, rest assured No, It Is The Drivers Who Are Wrong would be the interpretation.
You are quoting research that a graduate student of mine actually helped produce (not the original, but the followup in 1997), but you are missing the point. Yes, people will drive at speeds they are comfortable with, that is not in dispute. However, that is assuming there is no enforcement of existing limits. The research shows that if enforcement is factored in, most people will drive within five miles of the speed limit. It further shows, that this leads to a more consistent speed among all drivers on a given stretch of road and as such improves the number of vehicles per mile per hour per given stretch.
You are correct, posting speed limits does not affect overall speed. That is, unless those speed limits are actually enforced. If you aren't going to enforce them, then why post them.
It should be simple to understand that rules without consequences don't change behavior. OTOH, removing consequences tends to encourage the behavior the rule was meant to address, which is where we are today with speed limits.
The reality is that the faster vehicles go, the longer they take to stop or if the driver is distracted the further the vehicle travels until attention is focused. Couple that with the aging population, where reaction time is decreasing, and you have a recipe for disaster. One solution, albeit very costly, is to force people into driving autonomous vehicles. Of course, that will take years before they are ready and in sufficient quantity to make a difference (look at the hybrids, even government subsidies weren't enough to get sufficient people to buy them where they were economical and they were vastly cheaper than autonomous vehicles are expected to be). Or, a simpler solution, available today and proven to work is to enforce the traffic laws.
Again, I agree, just changing the number on the sign won't change peoples speed habits. However, enforcing the number on the sign has been proven to do so.
Which is a great theory, but the reality is that if the speed limit is set very low on a road for no apparent reason then a lot of drivers won't respect it,
Actually the reason is quite obvious. The speed limit is set unreasonably low so the police can pull over anybody they want for speeding. This might be to generate revenue (i.e. a speedtrap), or as a pretext to search the car for other more serious violations (i.e. drugs, cash, whatever). Since they can't legally just stop a car for 'looking suspicious' they just pull them over for speeding since they (and everybody else) are speeding.
Undertaking is illegal e.g. in the UK for a reason.
It's not illegal in the UK. I've gone past police cars on the inside.
This is also why cyclists aren't all being arrested, although to be fair in London it's also why several of them have died.
You cannot get gridlock in a traffic system where all rules are followed. The worst you could get is a potentially non-optimal throughput under special conditions. If you follow ALL of the rules, things will work out just fine. It's how the system is set up. However, I'm willing to be swayed if you can post an actual citiation and the report isn't bullshit itself, or extremely limited in scope.
Is it just my observation, or are there way too many stupid people in the world?
It should be equally safe for 2 cars to be both moving at 70 mph, or both at 30 mph.
That is demonstrably false. It is why we have speed limits in he first place. At the hgher speed the world is coming at you 40mph faster. That includes stationary objects and cars driving perpendicular to you. Your reaction speed and that of the other driver has to be much faster.
If the other car suddenly finds itself heading towards and debris on the road, it is far less likely that it will be able to stop in time. The only course of action might be to swerve or brake sharply. At 70mph you have much less chance of avoiding a collision, and if you get clipped and are sent towards a parked car or tree then the impact force will be much greater and more likely to be fatal.
In every way, 30mph is safer than 70mph.
Given various technologies like radar detectors and the Waze app, many drivers believe that they do have information about when it is being enforce.
Yes, and when the radar detector beeps, what do the drivers do? They slow down to the speed limit.
I concentrate far more at 120mph than I do at 60. Making me drive slower merely bores me, and adds distraction.
I agree the answer is autonomous vehicles. I look forward to those being available, even if it increases average journey times because everyone's going 10mph slower. Until then, I'm not going to do 40 in a 70 limit even if it gives me more reaction time because it's just not fucking safe.
It's obvious that our simplified mental model of how roads and drivers work bears little semblance to how they actually DO work, at least to anyone who has ever modeled any part of reality. To NOT have studied this, not to seen it as something which is properly the subject of a possibly long , expensive and anyway open-ended scientific investigation implies you're the (autisitc) type who mistakes systems of rules for reality, and perhaps enjoys doing so; someone who loves maps more than hiking. PRIOR to programming self driving cars, or at least expecting them to succeed and be marketable (maybe SDCs are a useful tool in this needing-to-be-done study of drivers and roads) you needed to know a whole lot more about how people make their cars behave in reality.
Here where I live, the opposite happens. Oddly enough, if you are on one of the feeder roads that is alongside a toll road, the lights are timed to turn red just as traffic approaches. At. Every. Single. Intersection. In fact, a former co-worker used to be a civil engineer, and showed that even with random chance, it could not have been designed worse.
The same civil engineer told me about the methodology in general. Congestion reduces wrecks, so it looks better for a city to have traffic moving 15-20 mph on the roads than at highway speeds because it lowers the fatal accident rate. There is no downside (to a town) to have streets that are impassible, and the local police can always make money by inspecting the license plates, looking for "busted taillights", and other things.
You're a selfish self-important cunt, although we knew this because you drive a Prius.
You're not the only person on the road. You have a social obligation to minimise inconvenience to others. Going at 45 in a 70 limit is just obnoxious; if everybody did that your fuel economy would drop.
Give me a week to observe the traffic pattern data, three independent drivers, and I will gridlock your highways without causing a single accident and, quite probably, by following the letter of the law. The problem is, generally, that everyone ignores the letter of the law. Those laws are put there for a reason (frequently) and it's usually (well, sometimes) a good reason.
Fortunately, after having grid-locked your traffic, I can show you how to fix it and, mostly, prevent it - though few ever listen to all the advice. That's why I'm retired.
"So long and thanks for all the fish."
The only reason why a robotic car would be unpredictable is that there are so few if them on the road right now that people aren't used to them. The reasoned, cautious AI driver will be entirely predictable. How many rear-end collisions at lights are because the driver behind assumed that the driver in front would go faster when approaching a yellow light? If people actually did what they were supposed to then the assumption would be that people would not break the law and everyone would drive like an AI car.
The great thing is that the accident rate for human-robot and human-human cars would both go down. Result! Impractical you say? Simply increase the consequences of being caught doing the wrong thing. Take away people's licences more often now. Too harsh? Then don't break the law.
Almost every single vehicle accident is caused by someone driving too fast for the conditions. On its outset, that sounds absurd. Think about it for a while (if you don't get it) and get back to me.
"So long and thanks for all the fish."
It's people who don't understand that simple rule that cause accidents
People who don't understand that the overtaking lane is also not an excuse for speeding nor does it have a requirement that you do 30km/h above the limit is also what causes accidents. In general idiots who think they own the road in every which way cause accidents.
I get that about once a week. Overtaking some speed limited truck takes about 10-15 seconds when I'm driving the limit. Some idiot will come up behind me at a stupid speed and start flashing his lights and honking the horn. They are just as bad as those people who block the lanes for no reason.
In the UK you can be pulled-over if you are driving in the middle lane and there is no one in the outer lane (left-lane in UK, right-lane in US).
You can be in the US too. And many states of laws that if you are not passing something within a certain distance (3 miles for Pennsylvania) then you better move over to the right lane.
Truth is like the sun. You can shut it out for a time, but it ain't goin' away. - Elvis Presley (source: imdb.com)
Err... No... The average idiot behind the wheel of an automobile is an inept idiot who is distracted and selfish. That is, by default, the assumption to make about every single vehicle on the road. There are a few things where this idea does work.
Here's a fun one...
Most areas don't use them. Want to ease congestion at a merge? Heh... Put up a "zipper merge" sign with a nice sign below it saying something like "Be alert, Zipper Merge ahead!" Don't go Full-Georgia. Let's just say that they've taken signage to a level that is above and beyond.
Anyhow, clear signage does help but the two areas you mention are not likely to help. Put lights on the signs to indicate when there are children present. Throw an empty cop car with lights flashing at the construction sites and do more work at night. Have a second rate of speed for when the school is not in session, by the way.
"So long and thanks for all the fish."
I've often marvelled at how many people will pass me on a two-lane hiway when I have the cruise control set at the speed limit. On a 30 minute commute, driving 10 or 15 Mph over the limit is going to get you there about 5 minutes sooner, so you'd be better served by just leaving on time. You can make up more time than that by just timing the traffic lights
Apocalypse Cancelled, Sorry, No Ticket Refunds
The actual problem with the speed limit rules is that the speed limits are everywhere (except when I drive on the Autobahn in Germany, in most cases where I take it there are almost no speed limits). So because the speed limits are everywhere people are not paying attention.
There shouldn't be speed limits in most places, only in places where the limits are truly necessary, but the reality is that all of this is built to generate income for the city, not to ensure safety or sanity of any kind.
You can't handle the truth.
So...everyone? The issue here is that people tend to drive as fast or as slow as the road allows, normally it's the common law speed limit. Humans can usually adjust to this, robots with strict rules can't.
That is incorrect. People tend to drive as fast or as slow as enforcement of the speed limit allows. If authorities start enforcing the speed limit, the speed driven will decrease. Since there is no real penalty to speeding, people speed.
Not quite.
There's been quite a few studies that have shown that roads have a psychology behind them, and that people will tend to drive more towards the speed limit that the road psychologically feels like it should be. You can make the speed limit whatever you want, but if people's natural instinct is that it should be faster then they will go faster. You can enforce a slower speed limit too, but that will just result in more tickets, not people going slower.
So if you ever wonder why there's an odd curve in an otherwise straight road...it's because they're trying to play to the psychology of the road to get people to naturally slow down. Fewer tickets, fewer accidents, fewer problems all around.
That said, this is something that's really only come up in the last 15-20 years so there are very few places that really take this into account.
Truth is like the sun. You can shut it out for a time, but it ain't goin' away. - Elvis Presley (source: imdb.com)
Erf... I should post this as an AC.
But, let me preface this by saying that I've been a professional driver, I've taken many advanced driving courses, I've worked in traffic modeling, engineering, and made quite a living at it. I also drive a BMW... I'll also add that I have absolutely zero at-fault accidents on my record in over 40 years of driving. This doesn't excuse my behavior but I hate highway driving. And I had a deadline...
So, I was tooling down I-95 corridor at 120 at times, just yesterday. I know, I know... However, it was not around traffic. Point being, you're correct. I really don't care if they fine me. They probably *should* fine me more for infractions.
Oddly, I have no infractions on my record but I did get a speeding ticket many years ago. Somehow, I've had none since. No, I have no idea why not. I do drive very differently when there's traffic present. They should probably figure out a way to account for the times when it is safe to speed but I doubt that will happen and I doubt that skill level will ever be taken into account.
"So long and thanks for all the fish."
If you are refering to the odd curve in the US interstate system, they are placed there on purpose to keep the driver alert. The concern when laying out the interstate system was that long straight highways would lead to fatigue and accidents, so the curves were inserted as an intentional feature.
With regards to driving as safe as people feel physically safe, that is false. Here is a test to try. Take a stretch of highway where people regularly drive over the speed limit. Now park a police car on the shoulder -- people will slow down, even though their physical safety has not changed. While one does take into account how safe they feel while driving, ie. slowing down in the rain. When talking about driving over the speed limit, it has been shown that lack of enforcement of the speed limit is the major factor. In short, it's not that people drive as fast as they think safe, but as fast as they think they can get away with.
A rule that makes no sense during congestion, I might add. That means its more of an exception than a rule in many places.
You just identified the real issue "following too closely behind the vehicle in front of you" perhaps if there was a legislative/punitive or technological solution to this, issues relating to "traffic waves" would be solved.
Well, we could just start giving out more tickets for tailgating. They're pretty rare these days except in particularly egregious cases, but tailgating (i.e., "not maintaining a reasonable and safe following distance) is illegal in a lot of places and already can incur a fine.
We have the technology -- police cars tend to have cameras mounted in them and if they are looking for speeding cars, they also can record the speed. Since "safe following distance" is usually related to speed, the speed + the camera can verify that the driving was unsafe.
The problem is that a "safe following distance" which would prevent traffic waves is a LOT bigger than most drivers seem to think, particularly when they're driving on highways at rush hour. If we just started ticketed everyone who didn't allow that distance, there would likely be huge public outcry.
Better tech solution -- with all of the car cameras and sensors these days, install a meter or dial which displays distance to the car in front divided by current speed, along with the speedometer or tachometer. The units could just be "seconds" to cover that distance. Then states can very clearly just say, "you need to allow an X number of seconds 'cushion' in front of you on highways at all times." And you can read that off the meter.
Then police would be completely justified in ticketing you if you had the meter available and chose to violate the law.
Good luck arguing for this, though. I think we'll be more likely to have roads full of autonomous cars before this would become common.
We need to ban human drivers as soon as possible!
That's what the pro-'autonomous car' people are saying right now. But that's never going to be the case and they need to accept that. What needs to happen is there needs to be reforms in the way drivers are educated, trained, and tested; the bar needs to be set higher, and testing needs to be performed on a more frequent basis. I cannot stress highly enough that the focus here shouldn't be on punishment; the focus needs to be on education and training -- but that having been said, incompetent drivers need to be excluded from operating motor vehicles entirely. Advocates of so-called 'autonomous vehicles' will now say 'Banning human drivers will solve all these problems', but the simple fact of the matter is, that all so-called 'autonomous vehicles' will always have a full set of manual controls and the ability to override the autonomous system without delay because, among other things, there will always be situations where only a human driver can get the job done; for this reason, even in a world where so-called 'autonomous vehicles' are ubiquitos, people will need to be better educated and trained at operating a motor vehicle, and tested more frequently to ensure that they're competent, due to the less frequent use of their driving skills ('use it or lose it').
Are YOU using the TOOL, or is the TOOL using YOU? Think about it!
I remember this debated in a criminal justice course, where the chance of getting caught would be less, but the penalties far harsher compared to a far less penalty, but a far higher chance of getting caught.
It was shown that having a far less penalty that was enforced heavily had a lot more impact. One example was the US drug laws where people could be ejected from college and perhaps go to prison for life for charges. As we can tell, it still isn't too tough to find weed.
On the other hand, various European countries have police will pull people over, ask for the fine cost on the spot, and let the driver on their own way after it is paid. Yes, the penalty is lower (especially without worry about points on a license), but people tend to stop misbehaving if they have to pay even a small fine, if they feel they will be caught.
Socrates figured people should ask questions about *everything*. We have a world where people get a glance at something, make up some stuff, and then stop asking questions about it and presume anyone who disagrees with them is a moron; on the other end, someone gets a certification telling us they know about something, so we stop asking questions and assume anyone who disagrees with *them* is a moron. Socrates wanted us to question ourselves *and* the experts, to find out why the experts disagree with us, and then to decide if the reason for disagreement was valid, or if we could even assess it with the information we had. Do that enough and you'll start realizing the experts are wrong--they're always wrong, just less-wrong than all of their predecessors.
Socrates lived in a time when the sum total of human knowledge was a little fraction of what we know today, and this it was a lot easier to get up to the state of the art in several fields. Sure, we should question experts, but on average, the risk of rationalising our snap judgement with half-understood facts and simplified sound bites is not insignificant. The expert is not always right, but he usually comes closer to the truth than a non-expert.
Really? Hmm...I'd beg to differ. Why? Well, there's quite a few things from that time that we still CANNOT figure out. For instance, how could the pyramids and numerous other structures so perfectly aligned (
No, there was an intelligence and knowledge that in many respect surpassed what we know today. Sure we have a good understanding of biology, engineering, etc; but much of what we know comes from the Greeks, Romans, and Chinese; much of which was been rediscovered. Probably about the only way in which we are more advanced is in our knowledge and use of high tech equipment to make CPUs and other stuff - but we do so using more rudimentary methods of making that stuff than the Romans and Greeks did.
IOW, there's very strong evidence that the knowledge in the ancient world was a lot greater than we have today. There was just a lot that was lost. After all, how long are our buildings and structures lasting for? There's Roman Roads that are still in service while ours have to be maintained every year; aquaducts that lasted as long as water flowed through them, and even then there's great sections of them that have been standing for centuries. The buildings today won't last that long - the materials break down too quickly.
Truth is like the sun. You can shut it out for a time, but it ain't goin' away. - Elvis Presley (source: imdb.com)
(I have worked in motor vehicle accident investigation, and am a mechanical engineer.)
At least this explains your irrationality, just like DEA imagines seeing drug dealers everywhere, you see every traffic situation as a potential accident.
When the inter-change of I696 and I75 was under re-construction, lane barriers were set up so that drivers had to commit to the lane they wanted 2 miles in advance. During that time throughput went up dramatically because the idiots couldn't lane change at the last second, when the barriers came down traffic went back to frequent gridlock at the interchange.
Michigan Lefts are work amazingly well when the traffic lights are timed propperly.
Apocalypse Cancelled, Sorry, No Ticket Refunds
There is always enforcement by peer. For example, Russian drivers almost always have dash cams. This usually means someone being a total dolt on the roads will be recorded by someone and that recording wind up on YouTube or played in front of a judge come trial time.
I saw a microcosm of this myself with my dash cam where the camera caught a vehicle popping into reverse, smashing into a car behind it, then the driver in front claiming it was the other person's fault. The footage I had on that MicroSD card likely saved someone's insurance premiums by a considerable margin because without it, it would be word against word.
I remember reading that dash cams are perhaps a definitive way to stop poor driving, just because once people get it in the head that their lane weaving and median driving might wind up on YouTube with their license plate available for all to see and look up, it likely would do more than any amount of police writing tickets for infractions.
You just identified the real issue "following too closely behind the vehicle in front of you" perhaps if there was a legislative/punitive or technological solution to this, issues relating to "traffic waves" would be solved.
Great sarcasm considering there is...
Truth is like the sun. You can shut it out for a time, but it ain't goin' away. - Elvis Presley (source: imdb.com)
The problem is that the programmers coded the "official" rules of the road. They forgot to code the unofficial rule. I call it rule zero, because it's the most important one. Avoid crashing at all cost.
All of the other rules of the road exist to supplement rule zero. They are meaningless if rule zero is violated. The computer code needs to reflect that.
One of our competitors trademarked the term "hypothesis". From now on, we will call them "boneheaded ideas".
If you're driving the speed limit, stay the fuck out of the left lane. It's not a complicated concept. If you need to drive slowly in order to drive safely, by all means do so, but that doesn't give you the right to become an obstacle to navigation. "In the flow of traffic" is always the safest speed, assuming your reactions and willingness to focus are up to it.
Socialism: a lie told by totalitarians and believed by fools.
This is the TCP vs UDP argument.
With TCP, you have flow control, and overhead, and you travel more slowly.
With UDP, you have faster data rates, but packets can be dropped.
Wouldn't you like to be "dropped" during your commute one morning?
These are my friends, See how they glisten. See this one shine, how he smiles in the light.
I wholeheartedly agree with Cederic on this. If you are practicing 'pulse and glide' anywhere but on an empty road with no other traffic around you, then you are inconsiderate asshole.
Ironically, pulse and glide would be acceptable to speed limit zealots, when it is very obvious this is very disruptive to other drivers and very likely lead to accident and unnecessary traffic snarl while other drivers try to navigate around such idiot.
Wrong, multiple studies by civil engineers have shown that designing and building a road and not posting a speed limit and then measuring how fast people drive (after a settle in time) and setting the speed limit at where the 85th percentile drives almost always establishes the same speed limit as what the engineers originally had in mind when designing the road. The only change to this is when conditions changes such as a volume increase because of additional residential areas or businesses being built.
Actually, here in Wisconsin the police are not allowed to write a ticket if you are driving within the measured 85th percentile speed (outside of school zones). The speed limits are measured automatically every so often on various stretches of roads by radar machines that report your speed to you along with a sign reminding you of the posted speed limit. This is to encourage people to slow down to the posted speed limit to keep the 85th percentile close to the prior posted limit.
It works really good and I believe other states follow this as well.
The most interesting thing that has happened as most states are raising their freeway / limited access speed limits to 70 or 75 MPH is that the 85th percentile really hasn't changed much. If a speed limit went from 65 to 70 the 85th percentile changes from 72 to 73 because people long ago determined where they are comfortable driving.
In Michigan I drive through a No Passing Zone that's also a School Speed Limit Zone and I still get passed while driving the posted limit. The School Speed Limit is 35MPH not the more typical 25MPH too.
Apocalypse Cancelled, Sorry, No Ticket Refunds
The proper answer, at least in the US, is to fix the laws so they reflect reality, and tighten up on the level of driving ability you need to have in order to get a licence, not just blame the entire population.
There is already a giant problem throughout the US of super low speed limits everywhere that in no way reflect the actual road conditions, capability of the road, or how fast most people already quite safely drive on it.
The stupid speed limits mostly came about because speeding has been abused as a great revenue stream, and special interest groups such as MADD have literally no clue but way too much power. In short none of this has anything to do with any actual road safety measurements or metrics.
The problem with setting unrealistically low limits is that they are very much increasing the risk of accidents not decreasing them, since there will always be a few of the most braindead people that have heard "speed kills" so drive significantly slower than the traffic all around them, cause massive bottlenecks, then self-righteously think they are being better drivers than everyone else. The truth that they apparently don't have the brains to grasp is that speed itself doesn't kill (otherwise just standing on the earth would be fatal), its speed differential (i.e. the thing they, and self-driving cars are choosing to cause) that is dangerous.
Germans teach to push breaks and if you can't avoid it, hit the animal, no swerving.
Random reward works. I'm not so sure you get the same results with random punishment.
Just another day in Paradise
If Red light cameras hold up in Court, I don't see why an unmarked car could set the cruise control at the speed limit and photograph every Idiot that passed him; He could work a section of road for a week or two before anybody would catch on. Some of the clowns I drive around could rack up enough points to loose their license, before the first ticket got mailed to them.
Apocalypse Cancelled, Sorry, No Ticket Refunds
More accidents than expected? Really? Any programmer worth their salt saw this coming a mile away...
The canonical example would be, an unnoticed speed change. The driver who does not intend to speed gets a large ticket in your system for making a mistake.
I'm saying that's a bad user experience.
I think the biggest thing we humans will learn from this foray into self-driving cars is that human behavior is much more complex than we have the ability to replicate artificially.
Autonomous vehicles definitely have a place on the roads, in commercial trucking. It's not as sexy and those Cali billionaires funding self-driving car research won't like it, but that's the real usable value in the tech. In 20 years, I can definitely see US highways teeming with caravans of self-driving, electric powered semi-trucks barrelling down a designated lane of the interstate.
Many aren't going to like it, but the simple fact is, driving a car around town is something illiterate humans can do, but programing a car to do it is not possible practically
Thank you Dave Raggett
Except that "everyone else is doing it" applies here. If I'm driving, and I'm conforming to the actions of the drivers around me, I'm not likely to get into an accident. If I can predict what the other drivers are likely to do (and it's a range of actions, sometimes stupid), I can drive more safely.
Now, throw in a robot vehicle that behaves like nothing else on the road and does things I can't predict from my knowledge of human drivers, and it will cause accidents. The smug asshole in the robot car can take pride in the fact that the robot was obeying the law and is not at fault, while leaving death and destruction in his or her wake.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
I'm going to say that the simpleton is the kindergarten dropout who can't read the fucking signs and came to a complete stop in the 4th lane 100 feet from the exit in the 1st lane.
If your exit is in the left lane then be in the left lane. If your exit is in the right lane then be in the right lane. If you don't know where your exit is then shut off your god damn phone and look at the fucking signs.
Speed limits are often not set to be realistic, but for political reasons. Human drivers usually drive the maximum safe speed on the road, which might well be what the traffic engineers would have set if it weren't for politics. This is not crazy driving. It's possible to drive at crazy high speed, of course, but that normally doesn't happen. Some drivers make stupid moves, but the drivers in general know what sort of stupid moves bad drivers make, and can handle it.
The current driving environment is created by people, and runs according to how people do it. Now, someone proposes to drop a vehicle in there that will behave nothing like a person, and blame the existing people for doing what people do.
The important thing in this world is people, not robots. Robots are things, and replaceable. If a robot messes people up, it's doing wrong even if it's legally in the right.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
The correct answer, of course, is to always prioritize the safety of the driver.
Not always, no.
For example, suppose you can hit a row of a lamp-post or a crowd at, say, 40 MPH. Now say the driver will *probably* survive either but it is safer for him to hit the crowd because all those human bodies he squishes will help slow the car down before he stops.
A responsible driver hits the lamp-post.
Could the roads handle nearly as much traffic if you rigorously followed the laws? If you had every car in America allow enough space in front of them to be space, you would have to drop at least 75% of the cars from every major highway in every metropolitan area in the country. The DC Beltway, the New England Thruway, Chicago, Seattle, Los Angeles...
I often see idiots like that in the RIGHT lane. Sometimes they get right on my butt even when they could get easily past me on the left!
Slow down, cowboy! It has been 4 hours since you last posted. You must wait another few hours.
I guess the philosophical difference is whether you want roads safer or you want rules followed more carefully. They are not the same thing. I certainly agree that draconian law enforcement will keep citizens in line. I guess the difference between us is I think punishment should be the last resort in changing public behavior. For instance, I would rather put more money towards funding inner city after school programs than simply hiring more police officers to reduce crime.
For the same reasons, finding ways to reduce vehicular injuries and property damage without relying on punishment should be the top priority. I admit this is an opinion of mine based on the type of society I wish to live in. I want a police force that is there to protect and not only to punish. And ignoring traffic engineering solutions to reduce traffic fatalities in favor of harsher enforcement of poorly designed laws is not a good idea, again in my opinion that is.
Less traffic laws make our roads safer. I find one paragraph quite insightful in this article:
Now to the silenced engineers and researchers. Federal law (Title 23) says fact–based sound engineering practices are to take precedence over conjecture. The problem, no one is willing to enforce it – including the FHWA.
Perhaps I am being hypocritical but here is a law I wish we did have better enforcement for. Instead of treating traffic safety as something we need to create more and harsher laws to engineer, how about we use more effective social engineering such as setting more reasonable speed limits or even (gasp) removing them from many roads all together.
-- All that is necessary for the triumph of evil is that good men do nothing. -- Edmund Burke
The problem with this argument is that it assumes there isn't an alternative. The alternative is to use a combination of technology and enforcement to change the behavior of the drivers. If there are reasons this can't be done (because the rules aren't reasonable), we should fix the rules. Nobody is suggesting that human drivers drive one way and autonomous vehicles drive another. The question is do we make the human drivers follow the rules or teach the autonomous vehicles to break them. The former used to be impossible. Now it's well within our grasp. There seems to be an assumption that you can't make the human drivers follow the rules but we certainly can do have the technology to do that now and should start to phase it in.
The driving algorithm for NYC if it followed that of people would be pretty easy.
1. Is there (almost) room for your car somewhere in front of you that is (mostly) on the road?
2. If yes, drive there!
It's not about lanes. It's about negative space.
It bugs me when a cop parks near the ramp I take onto the highway. I'm accelerating to highway speed when suddenly the guy ahead of me spots the cop and slams his brakes.
Slow down, cowboy! It has been 4 hours since you last posted. You must wait another few hours.
I agree that two cars doing 30 is safer than two cars doing 70. However, two cars doing 70 is still safer than one car doing 30 and the other doing 70 (at least assuming the 30 MPH car is ahead of the 70).
Slow down, cowboy! It has been 4 hours since you last posted. You must wait another few hours.
If there are cops everywhere then people slow down everywhere.
Good to see you have a realistic plan of attack here.
You are funny. At the start of my commute there are hundreds, no thousands of speeding cars. There aren't enough police in all of the state to ticket those drivers.
For some time now I have watches as driverless cars have been developed. And I believe that while the technology can be made to work, it will end up having a very hard time getting wide spread acceptance.
The primary problem is going to come the first time a driverless car is involved in a serious accident.
Since there is no driver, just passengers, how can they be liable for damages or deaths that are caused when the driverless car is involved in or causes an accident?
Does the liability fall back to the manufacturer of the car? Or the programmer that wrote the code for the car?
What company could accept the liability for all those cars they sold to the public?
Or would that be the issue, they can make the cars for $35,000 but the added liability coverage would be to expensive for most to afford.
Of course the real problem being discussed here is driverless cars sharing the road with idiot drivers doing stupid things everyday. They will be involved in accidents and they will cause major damage at somepoint.
The only way to solve that is to switch entirely to driverless fleets of cars instantly, and that is not something that can be done. Costs would be to much.
This can also be a very disruptive technology. I expect the first real commercial use for this would be long haul trucking. Setup a fleet of these driverless trucks and run them from depots on or near the interstates. They can run as long as they have fuel which should improve the transportation costs since the trucks don't have to stand down every 8 or 10 hours to allow the drivers to sleep. And being on limited access highways the chances of them being involved in situations that are unforeseen are minimized. Once at a depot loads can be shifted to normal trucks for delivery in towns.
Doesn't it just need also to be programmed to lie about it then? (And/or to "take the fifth"?) IOW, what kind of "legal intelligence" is also programmed in, if any? (I'm at least semi-serious here...)
If your brother's destroyed that many cars, maybe the problem isn't the "Asians".
We have a saying in Silicon Valley, DWA. It means "driving while Asian". We tend to have a lot of recent Asian arrivals who should not be behind the wheel, especially in the crowded freeways of the Bay Area. I was in an accident a few years ago where the Asian driver who barely spoke English panicked when trying to change lanes and hit a bunch of cars, mine included. I happened to see the whole thing since I was stopped in the exit ramp at the time.
Then again, bad drivers aren't limited to Asians. Like JMJimmy's brother in law, my sister has destroyed a lot of cars over the years as well, though the last time was clearly the other driver's fault.
This post is encrypted twice with ROT-13. Documenting or attempting to crack this encryption is illegal.
I don't know more about traffic than the experts. However, the listed practice is crowd-sourcing the speed limit, and crowd-sourcing is often better than expertise.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
Thing is, with unrealistic speed limits all over, drivers are conditioned to not obey the speed limit. If the speed limit was at a natural driving speed, the limit would suddenly become more reliable, and more people would follow it.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
In California, notably, you can be in the left lane all day long.
That's super helpful to know. I was wondering why traffic got so much worse as the number of people moving here from CA increased.
There used to be fairly good conformance to the "keep right except to pass" laws (even without the posted signs), but now the left lane is pretty much always packed bumper-to-bumper while the right-most lanes are empty. People enter the highway and immediately move to the far left lane, even if there's nobody in front of them and they're moving slower than the people currently in that lane.
Traffic, predictably, crawls now. I guess that makes the newly arrived feel more at home?
If you want a vision of the future, imagine a youtube comments section scrolling - forever.
You are wrong. Drivers naturally determine appropriate speeds for conditions and will drive those speeds. When speed limits are consistent with what the road allows, drivers do not naturally speed. In the US, speed limits are set unreasonably low to create a steady pool of speeders to draw from. That is not the case in other countries.
Driving as safely as possible is not defined as obeying the speed limit when the limit is unreasonable. Safely as possible, frankly, has little to do with arbitrary laws.
If you want to be as safe as possible, get off the road entirely.
Just because you run into a slower moving car doesn't mean you are driving too fast, there can be all sorts of reasons. Meanwhile, unusually low speed limits create speed differentials that are unsafe.
It's funny how so many people's logic on this topic, when carried to its natural conclusion, would result in a speed limit of 0.
"If you're driving the speed limit, stay the fuck out of the left lane. It's not a complicated concept."
Oh shut the fuck up. It's a speed limit, not a speed minimum. Speed minimum on most freeways where I live is 45. That you're a douchebag who thinks he has a right to go 10+ over the limit is also not a complicated concept. I see people like you all over Houston.
Do you look in your mirror, see him coming, and then change lanes in front of him anyway, knowing that you're going to obstruct the flow of traffic in that lane? If you planned your own maneuver better, could you have passed the truck in peace without causing anybody else to significantly slow down or having to deal with irate assholes? Because I drive a little over the speed limit and am regularly cut off by people deciding that they need to change lanes RIGHT NOW!!, when it means that I have to slam on by brakes to avoid rear-ending them.
Following the speed limit is not an excuse to break other laws, like "keep right/left except to pass" or obstructing the flow of traffic. Being antisocial and passive aggressive in traffic isn't morally superior to being that irate asshole.
If you want a vision of the future, imagine a youtube comments section scrolling - forever.
The problem is that the programmers coded the "official" rules of the road. They forgot to code the unofficial rule. I call it rule zero, because it's the most important one. Avoid crashing at all cost.
I don't know why you call this an unofficial rule and rule zero. In German traffic law, for example, it is absolutely official and rule #1: "1. Participation in road traffic requires care and consideration at all times. 2. Anyone participating in road traffic has to behave in such a way that nobody is damaged in any way, put into danger, or inconvenienced more than necessary under the circumstances. "
If we don't want road rage and frustrated drivers, we can fix the rules. If speed limits were normally reasonable, people would pay attention to them. I don't see any upside to strictly enforcing arbitrary rules.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
We always check who's driving when someone makes a bonehead move and nearly causes an accident. 5/10 Asian women (all ages), 2/10 Asian men (primarily late twenties trying to be Brian Tee on a budget), 1/10 Elderly Caucasian, 1/10 Young male (various races), 1/10 other.
if 220 feels safe, you've never experience a blow out at 160+.
There is no federal speed limit in the US
I'm still learning to drive, and have a problem like that. I want to stick to the letter of the law to practice for the road test which makes it hard to be on the road with people who are bending the rules. I have a decent idea of what I could do in everyday driving like go a bit over the speed limit, but I don't want to get into the habit before the test.
I listen to both RIAA and non-RIAA stuff if I like the music, tangential business/politics nonwithstanding.
Honestly, interesting and lots of food for thought. Been missing those on SD.
You lack imagination if you can't come up with such a scenario and you lack experience in communicating with humans if you can't parse a little hyperbole.
Luckily, there's still time to develop in these areas as you grow up. I think I hear your mom calling you to dinner.
If you want a vision of the future, imagine a youtube comments section scrolling - forever.
I think we can all agree on one thing though: If traffic laws are written on the assumption that they will be violated constantly in small ways, those are badly-written laws.
I agree. In that case the laws are probably written to increase ticket revenue instead of for safety. But drivers must have some discretion and some small leeway in certain areas of traffic law.
Most people, however, seem to think they (along with whoever they're cutting off) only need enough following distance to brake when the car immediately in front of them starts braking.
I don't really care for simply ticketing tailgaters. I'd much rather see them jailed, their licenses revoked, and their cars confiscated - a bit like what is done to extortionists, if and when the law catches up with them.
I've read a few things about this, and noticed that they don't seem to address a common problem: In high-traffic conditions, if you try to leave a "recommended" safe distance from the vehicle in front of you, what happens is that the drivers in adjacent lanes see that large gap, and fill it. Then you slow down to get far enough behind them, and that space gets filled in. Before long, you're driving at half the speed limit, people behind you are honking like crazy, and the ones that passing you are giving you the finger. ;-)
This is, of course, the opposite of tailgating. Some of us do try to keep a (relatively) safe distance, but other drivers make that impossible. So far, I haven't read of any solutions to this problem. Do you have a link to a solution?
(Yeah, I know; just don't drive in high-traffic conditions. That doesn't always work, either. Google traffic does help a lot now, but often there are no alternate routes that are any better. ;-)
Those who do study history are doomed to stand helplessly by while everyone else repeats it.
So you're saying the left lane is reserved only for illegal activities?
You're an idiot. The overtaking lane is perfectly valid to use when doing the speed limit, especially when other lanes are travelling below it. If you are driving 150 behind me and I'm doing the speed limit in the left line WHILE OVERTAKING, then my following words can not be stated enough: "Go fuck yourself you idiotic dangerous lunatic, I hope you get rear ended by someone equally stupid."
Do you look in your mirror, see him coming, and then change lanes in front of him anyway, knowing that you're going to obstruct the flow of traffic in that lane?
At no time did I mention I cut someone off.
Following the speed limit is not an excuse to break other laws, like "keep right/left except to pass" or obstructing the flow of traffic.
By legal definition you don't obstruct the flow of traffic if you slow someone down to the speed limit. And yes I fully agree if you're not going to overtake in a reasonable time don't do it. But don't expect a free pass if you're breaking the law in the first place by speeding. That lane is not your right, and if you can't follow a law (speed) then why should someone else follow a law (not cutting people off).
This is part of the whole not getting along thing that people are so great at doing on the highway.
*Sidenote: My pet peve is a 90km/h speed limited truck overtaking another 90km/h speed limited truck in a 100km/h zone. Not only is that actually impeding the flow of traffic, but when racing two misscalibrated speed limiters the result can take longer than minute and that just screws up the traffic fiercely.
Hlaf the problems with traffic are not the cars but the roads. Smart cars need smart roads, and mixed traffic ought to be separated as much as possible.
Label all autonomous cars with flashing lights and restrict them to AI lanes on crowded highways that have right turn exits only. Set AI cars to cruise bumper to bumper in phalanxes once they are on a freeway to save space. Allow phalanxes to travel faster than the regular speed limit. Eliminate contention intersections from those routes as far as possible. AI cars ought to communicate route info and traffic conditions in a network that encompasses their route and destination.
With smart route planning, AI designated routes, and community planning for this technology, AI cars will gain advantages in convenience and travel time over other cars..
You can't separate the vehicles from the system, and that system includes roads and other drivers.
Firstly, you don't have to enforce something 100% for it to be very effective. Even if the enforcement rate was 5%, that's still a 1 in 20 risk of getting caught, which is enough to put almost everybody off doing it.
Secondly, driverless cars are outfitted with a whole load of cameras and sensors. It would be relatively straightforward to save the most recent data in the event of a crash, which would make enforcing the laws trivial in the event of a crash.
Bogtha Bogtha Bogtha
About 12-13 years ago, as part of my Strong AI development work I did an analysis on autonomous vehicles as an application. One of things I came up with was a basic first step solution to dealing with the interactions between machine driven and human driven vehicles.
The basic solution is to always give a warning to other drivers that a car or truck is under machine control. My simple idea is a coloured light system (say green or blue) fitted on the roofs of autonomous vehicles that lights up to warn when the vehicle is under autonomous control..
Below the speed of light Special Relativity is one of the most accurate theories in physics - above the speed of light..
"Undertaking is illegal e.g. in the UK for a reason."
Contrary to popular belief, "undertaking" is not illegal in the UK.
It's _strongly_ discouraged and should be done with extreme caution. If you have any kind of crunch, bump or have the guy you're "undertaking" turn in on you as you pass, you _will_ be charged with careless driving, but undertaking itself isn't illegal (in fact, passing a lane hogger is accepted as a valid reason for doing it).
Unfortunately in the UK, the laws against being in the faster ("inner") lanes when not overtaking can only be enforced 1: By a cop and 2: when they've been observing for at least a few minutes to determine the culprit is a moving road hazard. the result is that the laws have made virtually no discernable difference to the prevalance of boneheaded selfish driving (invariably perpetrated by the "old men wearing hats" brigade).
Video evidence from 3rd parties(*), or police-operated road monitoring cameras isn't enough to file charges. The police have to apprehend the driver and physically ticket them, which is a traffic hazard in itself on a freeway and it feels like you're more likely to be hit by lightning than to see a cop writing a ticket for this behavour (only a few hundred tickets have ever been issued nationwide, whilst I see the behaviour 20-30 times per _hour_ on british freeways around London alone)
(*) If the video footage shows the driver actively preventing people trying to pass by changing lanes, that's a different matter and will result in dangerous driving charges.
"The primary problem is going to come the first time a driverless car is involved in a serious accident."
The moment the driverless car's video recorders show the human driver doing something boneheaded and illegal which caused the crash, the case will be thrown out of court - EVEN IF ALL THE OTHER HUMANS ARE DOING THE SAME BONEHEADED AND ILLEGAL THING. (caps intentional)
Passing slow/stationary objects on "no passing" lines is already catered to in most road rules (they usually state there must be 100 metres of clear visibility throughout the manouever) and being stuck behind a bicycle on a narrow road is no excuse for forcing your way past in a dangerous manner or crossing the no-passing lines (do that in a lot of countries and get videoed doing so, you can expect to be explaining to a judge exactly why you thought it was a good idea, as the guy who pulled that stunt and nearly crashed into me and 6 other oncoming cars on the other side of the blind bend he did it on found out - 3 month disqualification based on video from the riders and several of the cars which clearly placed him in the driver's seat)
Insurance companies usually work on a "knock for knock" basis - in short they know that in the case of a crash there is normally fault on both sides and normally no evidence other than "he said, she said(*)" so they eat the costs and agree not to claim off each other. When an automaton is shown to be "not at fault", this agreement is going to go out the window, especially when the prevalance of 360-degree video will uphold the issue.
An automaton which stops suddenly to avoid a kid running on the road and gets rear-ended by an inattentive following driver who didn't see the ball roll onto the street previously is in the right (as is a human. The whole point about following distance laws is to ensure the following car CAN stop if the one in stop does so suddenly for no apparent reason). That driver may well take it to court, but he's going to face a triple whammy of being in the wrong under current law (following too closely), picking up the other party's full legal costs when they countersue - because you can guarantee that insurance companies and manufacturers will do so to make examples out of humans clearly in the wrong or trying to make bogus whiplash claims, etc, AND probably finding that no insurance company will take his money afterwards, or if compelled to, insisting on seriously high premiums+deductables as he's demonstrated he's not only a bad driver, he refuses to accept being at-fault.
Just because humans do unsafe things by habit doesn't make it safe. Drunk driving being a case in point.
We've become less and less tolerant of this behaviour as it's become clear how much it really costs not just in direct costs but also all the knock-on expenses. Law enforcement and traffic management experts have been trying to force people to adopt safe following distances for decades, but my suspicion is that it will only start happening when freeway cameras recording roadspeed and following distances start issuing automated tickets (don't laugh, that's exactly what's planned on UK freeways from late 2016 along with 24*7 speed limit enforcement based on video evidence(****) instead of the current "random camera placement and only 90th percentile speeders". Law changes to hold the car owner automatically liable instead of having to positively identify the driver have already been passed.)
Google and Delphi have already tweaked their software such that it will break the speed limit on multilane roads if the surrounding traffic is doing so. That was explicitly mentioned in various articles as far back as 2008.
There is a fair degree of conflation/confusion in these comments of their low-speed electric vehicles (which are hard-limited to 25mph by CA law and mostly operating on 20-25mph roads(**)) and their IC-engine machines which have a lot more processing power and run at higher speeds.
(*) In car video evidence is mostly useless for crash evidence as it shows only a
In general freeways in most countries are designed to be safe up to about 100mph. Speed limits below this are fairly arbitrary although fuel consumption starts rising dramatically above about 55mph, which is why the USA had such a low national speed limit for decades following the 1973 oil crisis.
Humans are bloody awful at assessing risk, which is why tailgating is so common and why freeway pileups get very messy very quickly. It's an extremely low risk driving environment and people get complacent.
30-year studies have shown that people drive according to assessed risk. if the posted speed limit is felt to be too high or too low they'll gravitate back to that assessed speed and it's remarkably consistent across drivers. (too high/low: It was found to be if the posted limit was up to 10mph either side of the assessed speed people would stick with it but past that point they'd ignore it and stick with what felt safest.)
The problem is that "road safety measures" generally have the opposite effect to what is intended.
Narrowing lanes: People drive faster.
Parking restrictions: People drive faster.
More paint on the road: People drive faster
pedestrian guide fences: People drive faster
traffic islands: people drive faster
flush medians: people drive faster
Non-controlled pedestrian crossings: people drive faster
Light controlled pedestrian crossings: People drive even faster if the lights are green
What actually slowed drivers down was removing as much paint as possible and blurring the distinction between road and footways, or in extreme cases, cobbling the roads.
Case in point: London's Exhibition road https://www.rbkc.gov.uk/subsit... - although this hasn't worked nearly as well as the media tries to suggest.
Another case: Most of the Netherlands. People talk about the roads being pedestrian and cycyle friendly there but this only started happening since the 1970s. Cars are discouraged from rat-running on residental roads by making them one way, hard to manouever on and cobbled (which makes driving faster a noisy and increasingly scary experience especially when braking.)
People mostly drive badly because they don't want to drive at all. It's a chore, not a joy. In that environment they'll let a robot take over and be happier about the journey (you can do things during the trip rather than having it be X time stolen from your day)
More cops _on the roads_ would make the roads safer, full stop.
Assuming they're doing what they should be doing. Roscoe P. Coltrane and his ilk need not apply.
"people tend to drive as fast as the road and their skill level allows for."
People drive _well_ beyond their skill levels. They only find this out when they run out of road.
"I concentrate far more at 120mph than I do at 60"
You may well do, but those around you are not and being that much faster than the rest of the fleet on the road makes you a crash catalyst.
Similarly, stupidly slow drivers are also a crash catalyst. The issue is the spread of speed between faster and slowest drivers and ideally it should be less than 20mph, preferably less than 10.
It is at least as important to pick off the slow drivers as it is to take out the speeders.
"Then states can very clearly just say, "you need to allow an X number of seconds 'cushion' in front of you on highways at all times." And you can read that off the meter."
The usual figure is 2 seconds and lots of countries have signs up pointing to this along with "Keep right(or left) unless passing"
Unsurprisingly, these countries have lower crash rates.
"if you try to leave a "recommended" safe distance from the vehicle in front of you, what happens is that the drivers in adjacent lanes see that large gap, and fill it."
They wouldn't if they realised they were risking being ticketed for tailgating, or if better educated. The problem is that most drivers vastly underestimate safe following distances until such time as they find out the hard way.
"The speed limit is set unreasonably low"
Anything faster than 20mph in a residential area can be argued by statistics to be unreasonably high:
At 20mph there's 98% survivability in car vs pedestrian impacts
At 30mph that drops to 90%
At 35mph it's 50%
at 40mph it's under 10%
at 45mph it's under 1%
The old "E(k) = 0.5(mass) * (v^2)" equation...
Higher speed roads should be heavily segregated to keep pedestrians and other low speed users off and speeding enforcement in residential areas should be severe. 35mph vs 30mph being a good case in point about why
Roads have a design speed. The relationship between that and the legal speed limit is somewhat elastic.
I'm all for evidence-based policy making on safety issues, but in this case you do have to remember that those figures related to actual impact speeds in a collision, so arguing for setting the speed limit at those speeds is dubious for a couple of reasons.
Firstly, you have to consider that any normal driver is going to try to slow down or swerve to avoid a collision if they can, and so the relationship between the impact speed and the normal driving speed will depend greatly on the circumstances. Roads where drivers have few hazards to deal with and will have more warning of any potential collision can support higher normal driving speeds. Equally, on some roads with very limited room to manoeuvre and visibility, such as tight residential streets with cars parked down both sides and a single traffic lane between them, you'd be crazy to do as much as 20mph.
Secondly, you have to take into account the possibility of unintended consequences. Those figures only matter if an accident actually happens. If slowing the majority of drivers who will try to be reasonably careful down doesn't prevent an accident because they wouldn't have had one anyway at somewhat higher speeds, but it also prompts a small minority of completely irresponsible fools to overtake aggressively at much higher speeds and they are the ones who are involved in collisions, you might have increased the typical actual impact speed even if you reduced the mean driving speed on the road.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
The average idiot behind the wheel of an automobile is an inept idiot who is distracted and selfish.
If that were actually true, there would be far more collisions and far more damage/injury/death than there is.
You notice the idiots, of course, and for all the wrong reasons. If I drive into town to go shopping, maybe I notice four or five complete fools along the way. However, I probably don't pay much attention to hundreds or thousands of other drivers who I pass or interact with without incident during the same journey. I also probably don't remember as much the other drivers who went out of their way to be safe or helpful as a courtesy to help me on my own way, perhaps giving way to let me out at a junction, or driving carefully so that if I did make a mistake somewhere nothing bad happened because of it, even though there were probably at least as many of these particularly welcome drivers on the road as there were total idiots.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
I don't know why you'd reach that conclusion. Just because they're idiots doesn't mean they'll have a collision every time they mess up. Keep in mind that almost every single vehicle accident, ever, can be attributed to *someone* driving too fast for the conditions. (Not all accidents, of course. Just almost all of them.) Don't let them lure you into a false sense of security. If you didn't notice 'em that means they were just being lucky while you watched. If you were able to watch longer, you'd find 'em doing something stupid. It doesn't always mean there's an accident.
"So long and thanks for all the fish."
You've also got a lot of racists, too, apparently.
Keep in mind that almost every single vehicle accident, ever, can be attributed to *someone* driving too fast for the conditions.
That also seems like a dubious assertion. Some people certainly like to interpret any accident that could conceivably be speed related -- including a category in official reports that means there's no specific cause identified where excessive speed is one of the other possible choices -- as caused by speeding. Others might argue, reasonably enough but not particularly insightfully, that any speed at all is probably too fast if you're stupid enough to drive while drunk, drugged, tired, on the phone, or otherwise clearly impaired. If a single-vehicle accident doesn't include collisions between a vehicle and some inanimate object that were a result of evasive action when someone not in a vehicle did something dumb, like stepping out into the road while drunk or riding a bike out from a side road without looking, that probably helps too. But all of these common arguments seem like convenient retrospective positions that distort the real causes of accidents to make a predetermined point about speed.
Now, if you wanted to argue that most single-vehicle collisions were caused by the driver screwing up badly one way or another, that would be a much more reasonable point, IMHO. If you're driving competently, then in the absence of some sort of mechanical failure or some external influence that causes but isn't caught up in a crash, it's pretty hard to cause a collision. But again, since single-vehicle collisions are quite rare, I'm still not sure this supports your original position that the average driver is an inept idiot.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
I didn't say speeding. I said driving too fast for the conditions. If you were unable to stop in time, then you were driving too fast for the conditions - not necessarily speeding. If you're driving five miles per hour down a snowy lane and you slide into the ditch, you were driving too fast for the conditions. Sometimes, and in some vehicles, that speed is any speed greater than zero.
"So long and thanks for all the fish."
Yes, I understand the distinction. I was just trying to be correct, since in the particular case I was referring to, it is speeding -- as in, exceeding the posted limit -- that often seems to be cited. This is typically then used in an attempt to justify something about speed limits, when as you quite correctly point out, it is whether the speed is appropriate for the conditions that really matters.
However, please note that I was only referring to speeding in one specific part of my post. I think the general point that unless you go to the extreme of saying (correctly but not particularly helpfully) that any speed is too fast if a collision results) there are a lot of other factors that can and do also contribute to single-vehicle collisions. If someone has twice the legal blood alcohol limit in their system and then collides with a tree while trying to manoeuvre at 5mph in the pub car park, I think most of us would probably agree that the impairment due to alcohol was the main cause of the accident, not the fact that the car was moving at an otherwise normal speed but beyond the control of that driver at that time.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
To that I'd respond that they were driving too fast for the conditions, still, and that alcohol was certainly a contributing factor. The conditions are not necessarily road conditions but the condition of the driver is included in that. It may not seem particularly helpful as a metric but it is - it's the root cause. If we can solve that, instead of treating symptoms, we can pretty much eliminate traffic accidents or even vehicular accidents. Unfortunately, people are really not that good at driving and aren't nearly as skilled at it as they think they are.
I do have a kind of unique perspective. I worked in traffic modeling for many years and was a professional driver (trained in the military and that was my job while I served for quite some time). Since then, I've also become quite an automobile aficionado and I've taken many, many advanced on/off road and specialized driving courses up to and including open wheel, on track, training courses. If you crash, you were going too fast for the conditions. If multiple people crash then chances are that multiple people were going too fast for the conditions. Obviously, if it were a race, the objective is to find the edge of that envelope and stay within the boundaries as close to the limits as possible.
I guess I could see someone seeing it as not a useful metric but I humbly disagree. I think it's the most important safety element to keep in mind. I can also say, with some authority, that few municipalities actually seem to be all that keen on taking the advice about speed limits. They paid, quite handsomely, for the advice but they don't seem too keen on taking it.
There are many times and places where it would be perfectly safe to exceed the posted speed limit (assuming it is without traffic or that traffic is all doing that same or similar speed). That, to me, means the speed limit should be increased but, at the same time, people need to be aware that the speed limit is the maximum speed at which it is safe to travel with ideal conditions. Rather than educate drivers they'd just as soon keep the speed limits arbitrarily low. Speed, in and of itself, isn't the cause of the majority of accidents. Speed *and* conditions are the cause - almost invariably. Sometimes any speed greater than 0 MPH is too fast for the conditions, like when you're drunk and hitting a tree at 5 MPH.
I'm pretty sure we're saying the same things, for the most part, but differently. I've tried to articulate it a bit more clearly.
"So long and thanks for all the fish."
"but it also prompts a small minority of completely irresponsible fools to overtake aggressively at much higher speeds"
Such fools will rapidly find that noone will insure them in an automated world.
Crackdowns on this kind of driving are already a thing across a lot of the world, helped immeasurably by the proliferation of dashcams.
And yet in my heavily cyclist-populated city here in England, that small minority of car drivers still manage to cause damage and sometimes even death, so whatever "crackdowns" you're talking about certainly aren't happening effectively here. :-(
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
Yes, it seems like we agree on the fundamentals here.
I sometimes wonder whether having formal speed limits everywhere actually does more harm than good. My experience has been that most drivers will try to pick a reasonable speed for the conditions, and often be much better at it than some blanket policy set 20 years ago in an office about what the default speed limit for a certain type of road should be. However, this gets thrown away in two circumstances: either the driver is irresponsible (impaired, just doesn't care, etc.) or the driver doesn't pick up on something that would indicate a lower speed (still their responsibility, but qualitatively different in that they would do something about it if they were more skilled/aware).
The first group probably ignore speed limits much of the time anyway and hope not to get caught, but the latter group may be prone to seeing a fixed legal speed limit as a target. We all know it shouldn't be, but at the same time, I think it's fair to say that almost everyone hates the guy who's doing 15mph under the limit and holding up a whole queue of traffic when there really is no good reason to do it and nowhere to pass safely. I wonder whether we wouldn't do better to have some sort of advisory range of speeds in most places, perhaps with some sort of presumption in law that if you choose to go faster but are then involved in an incident then you will bear some of the responsibility. Then we could reserve fixed, prominently signed speed limits for specific places with specifically stated reasons, and actively enforce them with high visibility and substantial penalties for violation.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
What you are calling arbitrary rules are very carefully decided with the input of a cadre of civil engineers as well as the local communities and based on a consensus. They are no more arbitrary than the fact that we drive on the right-hand side of the road. If you think that something has gone wrong setting the limits, you can get them changed through various processes. Many states have recently increased their limits. This argument is often put forward and falls down the same way. If you're a qualified civil engineer who disagrees with the civil engineering work and want to offer you services for free to a community, they're probably willing to listen. I know that my county is paying a lot of money right now for a traffic study in the area where a school may be expanded. Too bad you didn't bid a lower price.
This. If people would, say, lose their license for a year after a third penalty within a set amount of time (say, 1 year), they will comply. If they just have to pay even a high fine each time, it's not a deterrent. The penalty for continued noncompliance must be very harsh, to ensure compliance. Losing your driving privileges (not rights!) for a significant period of time is most likely harsh enough to make a difference.
And, before someone counters, I don't think it will be ignored, like DUI penalties are ignored many times. Simply for this reason: your judgment is impaired when you're cited for DUI (obviously). However, the majority of speeding is done during lucid, non-impaired times, so the speeder does indeed understand the penalty and outcome consciously.
You miss my point. I respect the traffic engineers. However, once they've said their piece the speed limit is then set by a political process, and I don't have respect for that.
Right now, we have a remedy for too-slow speed limits: we speed, and generally drive at the speeds the engineers would recommend. If we had to obey them, we'd be dumping an informal process that works fairly well for a largely untested formal process. If we had to obey the speed limits, we'd have to go through the political process to get them to what the traffic engineers say, and that's slow and uncertain.
Back in the 60s and 70s, we were all pretty sure that marijuana would be legal once we grew up and started electing people who thought as we did about it. Currently, it's still a Schedule I drug federally, and a few states allow it anyway.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
A political process where the maximum speed that can be set is usually the maximum speed determined by the civil engineers but communities have the option to increase safety by setting limits even lower. That seems a pretty reasonable thing to be left up to local control. The only other logical conclusion is that everything should be set by the less accountable federal government including things like zoning rules.
Anyone who thought that automatic cars would have less accidents, is very naive.
Or, perhaps supersticious. As in: "Computers are Magic and therefore Perfect!"
There is some evidence to support the idea so you're guess is correct. On closed highways, some of them, we have both a minimum and a maximum speed. Those are for ideal situations. They have emergency signs that turn on and show a new limit during snowy/icy conditions. During that time, there's no minimum speed limit but - theoretically, if traffic falls below a certain speed they're to close the highway because it's likely due to visibility or inability to control the vehicle's deceleration speed.
Unfortunately, this varies per State with no real limits any more so you get people from outside the area (America's kind of broken) who aren't familiar with icy roads that are covered in a layer of snow who then cause large accidents or, more amusingly, end up spinning around and around until they finally slide off into the ditch. It's actually not that often that they get hurt so it's okay to be amused. It really is quite comical to watch. They're thrashing about frantically all while the car is being quite graceful and going where it was going to go no matter what you did. Of course, if they'd stop thrashing about they had a chance at recovering from the skid.
Anyhow, some areas have multiple speeds for multiple conditions (such as trucks are limited to a slower speed or not allowed at all in certain weather). This is backed by pretty sound research. It's not universal and it is often ignored but the idea is sound. Unbiased and frequent enforcement of traffic laws does actually help but speed traps aren't usually a good idea. If you're really curious, I can dig out a few papers as I'm pretty sure I still have them stored on my home network and I can access that from here.
"So long and thanks for all the fish."
Has it occured to anyone that the problem is not with the people or the cars, but with the laws?
The majority of our traffic laws and their limits are designed to be broken on a regular basis by everyone.
They are designed to be revenue generators, not to keep the roads safe.
The research I've seen here in the UK supports similar conclusions. Unfortunately, at least for now, the only major variations we have in default speed limits are for different vehicle types.
We do now have variable speed limits in place on some of our motorways (our most significant highways) where in response to volume of traffic, or sometimes because of some other incident or poor environmental conditions, a lower limit than the default 70mph can be imposed and is displayed very obviously on digital signs over or alongside the road, many of which also carry cameras for enforcement purposes. These are one of the few changes in our road traffic rules in recent years that seems to have been a clear win, because as you might expect they do seem to promote smoother traffic flows in congested conditions.
Unfortunately, that system only applies on specially prepared sections of our most important roads, so it still won't allow any discretion to drive a bit faster on a literally deserted road on Christmas morning, nor lower the ::ahem:: target speed ::ahem:: when, say, there's a bank of fog just over the next hill on a rural road.
If you disagree, post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like.
Canada has a strange system as I understand it. It's rather subjective and, of course, has all the potential trappings that go with subjective enforcement. I may be mistaken but I'm pretty sure I am not as I've done some work there, although it was not me personally but a few employees who did the work. I should also probably know this for certain (I don't) because I've got dual citizenship by grace of heritage. (I'm mostly Micmac.)
Now, in Canada you can, theoretically, go as fast as you want (in certain areas) as long as you're below the maximum speed limit. However, you can still be going slower than that speed and be cited for something akin to "unsafe driving speeds." Theoretically, I could be going along at the speed limit and you could be going along at a slower speed. I, with more training, can be doing fine and handling my vehicle in a safe manner while you may be stopped if the officer thinks you're driving too fast and exhibiting signs that you're going too fast for the conditions.
And, for amusement: Driving in the snow is dangerous but, eventually, you end up with giant snowbanks. I own a number of vehicles and not all of them are meant to go to shows or the likes - some are meant for fun. I have a 1978 Datsun B210 that has only one goal in life and that's to serve me in absurd ways. I do almost nothing with it that it was meant to do. I will drive it through the woods and to the top of mountains. It has been rolled over multiple times and not by accident. There are but two seats and 5 point restraint harnesses and a helmet should be worn.
Anyhow, there's a steep hill near my house (a lot of them, actually) and I'll crest the hill, make sure nobody is coming, lock the e-brake, and cut the wheel quickly in one direction or the other. We then 'pinball' down to the bottom of the hill, bouncing off of snowbanks, and usually with very little control or no control at all. You just let up on the e-brake and apply some acceleration, and try to keep it going all the way to the bottom should you end up being close to stuck in a snowbank. It's a standard shift so it's a whole lot of work and a buttload of fun.
It's even reasonably safe. I live in an unincorporated township that has six residencies and there are times of the day when the odds of traffic are near zero, plus you can see if anyone's coming from a long ways away. The snowbanks are as high as 10 meters and rather soft unless there was a thaw and freeze spell just prior.
Well, not just for amusement, I guess the point is that safe, conditions, and acceptable risks are all very subjective. If you ever want to try it, let me know when you're in the States and I'll even let you drive it. Another fun one is to go roaring into town at night, hit the e-brake just before you hit the snowbank, you hit the snowbank sideways and flip the car upside down and go skidding across the ice. Then you get out, flip it right-side-up and wait for the oil to drain back down and you're good to go and do it again. You might as well visit the old guys out ice fishing while you're there. They've come to expect me to behave a bit like this and I've even managed to get a few of 'em to ride with me.
I used to want to be either a stunt-driver or the guy that drove the car for television commercials. That's why I chose motor-pool driver for my first selection MOS, why I was curious about traffic in uni, and why I've since taken many driving courses. I have zero at-fault accidents on my record. The examples above were not accidents, they were on purpose. ;-)
So, it's very subjective and we'd need intelligent and unbiased enforcement if we want to go with a different system (I think). I suspect we'd be far better for it *assuming* that we had well trained enforcement officers. Maybe we should look at Japan's police training or the training given by the Alaska State Troopers? The RCMP isn't too bad and might be worth looking into.
Sorry for the novella but I figured I'd fill in some info and give some examples. For science!
"So long and thanks for all the fish."
If you are refering to the odd curve in the US interstate system, they are placed there on purpose to keep the driver alert. The concern when laying out the interstate system was that long straight highways would lead to fatigue and accidents, so the curves were inserted as an intentional feature.
While much of the discussion has been around the Interstate system, I'm not referring to simply the Interstate System, but all roads. And per the "odd curve" I'm referring to things like side-roads having a small curve in them (
With regards to driving as safe as people feel physically safe, that is false. Here is a test to try. Take a stretch of highway where people regularly drive over the speed limit. Now park a police car on the shoulder -- people will slow down, even though their physical safety has not changed. While one does take into account how safe they feel while driving, ie. slowing down in the rain. When talking about driving over the speed limit, it has been shown that lack of enforcement of the speed limit is the major factor. In short, it's not that people drive as fast as they think safe, but as fast as they think they can get away with.
Presence of an officer may have a small impact; however, that will only be for the short distance around the officer. Left alone, people will typically follow what the psychology of the road itself will mentally suggest to them. This is based on the curves, stop signs, traffic lights, distances between such things, how many lanes there are, etc.
For instance, adding a second lane each way bumps up the speed that people think they should be going - enforcement of the limit doesn't change that psychological effect; it's better to take the psychological effect into account when designing the road or changes to it as you can actually make everyone slow down (or speed up) regardless of the presence of an officer.
Truth is like the sun. You can shut it out for a time, but it ain't goin' away. - Elvis Presley (source: imdb.com)
It is well known that 90% of people think (incorrectly) that they are a better than average driver. Automatically needing to re-sit AND re-pass your driving test on a regular basis may help to disabuse people of this erroneous notion. Also, since driving conditions and laws do change on a regular basis, people's tuition about driving and skill sets do need to be changed on a regular basis. And we know by experience that people will not undertake that tuition voluntarily.
Birds are not dinosaur descendants;birds are dinosaurs, for all useful meanings of "birds", "are" and "dinosaurs"
I've been thinking about why the Google cars are involved in more (minor) accidents than non-google cars. Based on no actual knowledge or facts I think the reason may be that the google cars are driving too well. The google car has much better situational awareness than a human driver and will therefore brake in order to avoid potential situations that no human driver would even be aware of. This unexpected braking is surprising the human drivers and they are running into the back of the google cars.
The solution would be to lower the threshold of when the google cars should brake to avoid potential situations -- in other words drive more like a person, i.e. more dangerously.
That or just expect the humans to adjust their driving to accommodate the safer google cars as they get used to them being on the road.
"Grab them by the pussy" -- President of the United States of America