Yes, this is why any statistics about driving that are meant to be compared across cultures need to have "distance driven" somewhere in the denominator.
Yea, I don't know I disagree. The denominator should always be just be population. I don't want to use sprawl as some sort of justification for more deaths per capita. Sprawl should be reduced and/or mitigated somehow.
I agree with the parent that distance driven should be in the equation. If there's a certain risk percentage associated with driving as a whole, then it's almost certainly somehow bound to distance driven.
For example, the risk of a car crash should be near zero if the distance driven is also zero. But even in the safest of vehicles, driving 100KM, 1,000KM, or 10,000KM+ each are going to have higher rates of accidents assuming real-life road conditions.
In the US, on average, people need to travel further to get to their destinations, and so on average I'd expect car accident and death statistics to likewise be higher.
I'm very in favor of reducing urban sprawl, utilizing public transit, etc., but in present state the US has vast quantities of existing road networks which can't be consolidated or mitigated overnight, so the next best way to improve those statistics would be improving traffic controls, vehicles, safety features, etc.
Measuring the wrong thing results in the wrong outcome.
Many of the most effective and efficient measures to reduce traffic fatalities work by reducing traffic. For example, promoting public transit. However, if you measure fatalities per kilometer these highly effective mechanisms have no effect on the metric.
It depends on which level of government we're talking about. The city governments can only really improve traffic controls.
But state and federal governments should absolute zero-in on reducing sprawl and travel distances and those actions should've started yesterday and they should stay laser focussed on reducing driving distances (especially for non-leisure reasons)
Different metrics measure different things. The right metric depends on what you want to compare and what kinds of claims you want to make, which determines what should be adjusted for in the metric.
EU has much higher training and licensing requirements in general.
In the US a 15 year old can get a learner's permit and start driving (with an adult) the same day. They can be licensed to drive on their own at 16 by passing a fairly cursory written exam and a short road test. No formal/classroom instruction is required.
Raising requirements in the US is prohibitive due to the lack of options individuals have when they cannot drive. A 16 year old often needs, or wants to work for an income. Owning a car is often a job requirement due to the distances involved in american cities. Getting a job within walking distance is often a non-option due to availability of employers, and biking is of limited safety profile depending on the area. If the only way to get to your job is through a state highway - driving may be the only safe method.
Lisence requirements vary by state. Here in Maryland, you need 30 classroom hours and 60 hours behind the wheel before before getting a provisional license.
Here in the NL I'd say it's at least €3600 if you have zero experience. This is my estimation for both theory & practical parts based on my own experience, current rates, and what little statistics I could find. Often much more if you fail and have to take more lessons.
This is speculation on my part: I assume US citizens tend to drive more short distances within cities to get to places, like shopping, while many EU cities are walk-able.
Cities and city entrances have the largest concentration of people and accidents.
I can't speak to every European country, but Portugal also has a lot fewer stop signs and traffic lights than the US (roundabouts are one reason, but there are multiple four-way intersections on my street that would have stop signs in the US).
Given the way American streets are set up, rigorously enforcing stop signs is probably beneficial, but I think other factors about how streets are arranged and how people drive are more important.
Possible causes for this include the prevalence of really large SUVs, which make it physically much more difficult to even see pedestrians - especially children.
Another part is truck design. Same reason: American trucks have elongated noses for the engine, whereas European trucks have the driver sitting directly above the engine.
On top of that, European countries have much more strict testing requirements on vehicles. Basically, every 2-4 years you have to have your vehicle inspected for roadworthiness - foundational stuff such as structural rust, worn-down tires or brakes gets caught much, much earlier than in the US.
To be fair to the US, various states have yearly inspections that include undercarriage rusting and other issues. It's a factor but I think less important than the truck part.
Fatalities are horrible to the people involved, and in many cases the people responsible for such are punished. This is about saving a fractional percent of people while backdooring the technologies as a surrogate for control of everyone else.
The error in agreeing to automatic enforcement lay in the indirect failures that naturally follow within centrally structured systems, when those automatic systems stop working correctly, or worse selectively work; the world will be worse off than not having the solution in place at all.
There are dramatically more risks of this becoming a component of a panopticon prison in a fascist state, something the US is degrading into right now with the slow erosion; and stress fractures to our rule of law, it might very well suddenly fail overnight.
What impact will these solutions have in breeding discontent if everyone has the boot of the government on their neck every time they roll through an empty stop-sign where no one is there..., or worse when they did stop and the AI mis-categorized it as a rolling stop. What feedback systems correct a faulty running system? Government and government apparatus have trouble getting sufficient benefits to legitimate welfare recipients, what makes you think they'll do this any better? Competency is not a common trait for government workers.
Who do you think will be most impacted, the people with less awareness, or the people with more awareness. Lower IQ/cognitive speed vs. Higher IQ/cognitive speed. Would this result in an evolutionary filter against intelligence?
Would these dramatic changes drive the intelligent people which society rests upon (dependently so), so crazy that they end themselves, don't have children, or end their children and themselves? Is there any hope for a future under such repressive and stagnant systems. No there isn't. Intelligent thought is largely based in cognitive speed, and multiplied by education. There are some very educated people who are not necessarily sufficiently intelligent to stand in for these people. Their words and ideas often cause more harm, the more complex the system becomes.
The moment you rest an argument on do it for the dead people, or do it for the children, which is what %, you dismiss all the failures of the proposed system. Those failures still occur, those harms still happen, and the type of people you have left are less capable of adapting, or rather become enraged with each additional reminder that they are not people, they are slaves or animals.
A nation becomes strong only as a result of its strong people in unity.
When you make people necessarily dependent on the imagined detriment of what could happen, prevent them from acting, and do this at the expense of what is actually happening, you get a weak fragile complacent brittle people who break and are parasitically dependent on a pool of people that shrinks to nothing.
These detrimental characteristics spread over time both laterally among people but also generationally, and eventually circumstances occur where your people simply cannot adapt to what the environment requires as needed, and in that existential threat you face oblivion as a species, extinction.
Complacency, and a blindness or reactance to the risks, breed delusion which takes root spreading to those that remain, as a contagion.
The moment you think you can make people better by treating them like animals or slaves, or prisoners, is the inflection point towards your people's ultimate destruction; although it may be many years between. Every person is dependent on every other person indirectly, and some carry more than others.
How do you suppose such camera's of an all seeing eye will change the populace for the worse, might it make them more animalistic, ugly, violent... just as Sauron did as described by Tolkien, and much of the basis for Tolkien's works is based on the bible.
The only way to win a game of thermonuclear war is to not play the game. The same can be said about a lot of decisions which pigeonhole your future into a box without a future.
Police no longer feel the need to do their jobs, and Americans in general have just lost any sense of empathy or even awareness of other people.
But also we have a serious problem where taking away someone's license to drive is to sentence them to poverty if not homelessness and starvation. We don't have decent public transit and there are very few jobs within walking distance of most residential areas. Those jobs that do exist don't pay a living wage because pegging minimum wage to inflation or even the poverty line is "communism" and an "attack on businesses".
Our problems with car fatalities is really only one small symptom of the ongoing collapse of American society.
My objection to raising minimum wage too high is that it makes it illegal for people with modest skills and abilities to earn a living unless they can find an employer to dress up charity as a job.
I know people who would struggle to create $100 of value for an employer per day. I would rather they be allowed to hold a job at $80/day than to have minimum wage set above the value they are able to create.
Nothing wrong with a "rolling stop" if the sight lines are good and there are no pedestrians or crossing traffic. The point of a stop is to allow traffic to cross the intersection in a safe and orderly fashion. If you slow down, and verify that everything is clear, then that objective is achieved even if you don't come to a complete stop.
If these cameras were smart enough to issue citations when pedestrians or cross-traffic is present I could support it. But issuing a citation at a deserted intersection when no risk is created is just absurd.
Then cities should adopt the yield signs that say that.
I agree our system could be much smarter. So may timer-driven systems would be better if the computer knew the presence and number of cars, pedestrians, bikes, etc.
I could support this if you combined it with criminal and civil liability when you guess wrong and run someone over while blowing your stop-sign. Right now, that's a $500 ticket at best, and it happens every day.
The whole problem is that people don't look for pedestrians -- they look for another car that might hit them. So they are looking the wrong way. And then they tell the cops some sob-story about how the dead pedestrian "came out of nowhere".
Um, no. If you hit and injure someone in a crosswalk you can be sure you will face civil liability. Hope you have enough insurance (the legally required minimums are far too low).
> If you slow down, and verify that everything is clear, then that objective is achieved even if you don't come to a complete stop.
There are too many failure points there to trust mediocre meat sacks to follow that process correctly. Remember that driving rules and restrictions are not written assuming an alert, effective, and skilled driver operating a well maintained vehicle, they are written assuming an average person who has successfully completed a driver's test driving something that passes basic road worthiness checks.
It works well enough in other countries than America.
E.g. in the Netherlands or Germany there's no need for 4-way stops and such. If no other signage applies, then whoever is on the "right" side has priority over the people on the "left" side. And exceptions do apply, i.e. it's not "that simple" either. It does depend on whether both roads are on the same level or not. A road to your right that has a sidewalk border stone running across it does not give them the right of way, while if the sidewalk border stops and both roads intersect directly at the same level, then the road to your right does have right of way.
So e.g. if you take a typical urban development with lots of little streets and houses, where you'd see a lot of rolling stops in America, nobody's gonna stop at every intersection, rolling or not there. This does go as far as when cars from all directions arrive at the same time, then nobody has automatic right of way and one of them has to wave the person to their right through and will be the last allowed to proceed.
It's about the incremental injury / death; full speed is going to cause more. Going through at 3mph shouldn't cause more, but if rolling statistically does then full stop should be enforced.
But what is the point of rolling through at 3mph? To save 2 seconds? The reality is people roll stops because they are a bit lazy, and stopping fully requires marginally more effort. And while that laziness is harmless at empty intersections, it invariably turns into complacency and habit that bleeds into situations that aren't as safe. The same dynamic happens with signaling lane changes.
The problem is that you can't always tell that there are no pedestrians or crossing traffic unless you take enough time to come to a complete stop. There are plenty of stop signs where that's not true, but also plenty where it is, and it's not always clear which one it is to the driver.
I think the right fix here isn't total enforcement nor relaxed enforcement, but relaxed signage. If sight lines are good enough that it's safe to roll through, that should be a yield, not a stop. Stop should mean, you actually need to come to a complete stop to safely navigate this intersection. Then you can enforce it without qualms.
Good point, another thing that the EU tends to do differently from the US. US rarely uses "Yield." The only places I generally see it are at roundabout entries.
Often on EU roads they will use "sharks teeth" yeild markers where a side road enters or crosses a main road. The requirement there is essentially "proceed if clear" a full stop is not required. I have rarely (maybe never) seen a US-style 4-way stop there (my experience is limited to Scandinavia and Germany).
Signage is ineffective in addressing short term environmental or visibility impacts. Sure, it might be easy to see during the day with clear visibility. What about at night? Fog? Snow or rainstorm that is restricting visibility? Some dropped a storage pod on the road that obstructs the view of everything except a yield sign?
Definitely not, there are tons of stop signs in places with low speeds and great sight lines that are perfectly safe to treat as a yield. And conversely, there are occasionally some interesting places (on-ramps for very old freeways) that have a yield where it's not safe to proceed without stopping first.
Nah, spoken like someone repeatedly humbled by the complexity and detail of domains other than my own.
“I can drive a car therefore I understand traffic design better than traffic designers” is obviously an absurd statement when you just say it outright instead of condescendingly implying it.
> HN commentator revolutionizes traffic design with groundbreaking insight: consider the sight lines
With computer vision the case of checking for pedestrians in the vicinity is trivial. So these cameras are definitely worth it for that
I do disagree about the rolling stop though. After drunk driving, drivers getting too relaxed and working off of predictive execution has to be the biggest cause of road accidents. A driver rolling past a stop at high speed in a school zone cant react fast enough to kids running past or even just walking on predictive execution themselves because they think the car will stop.
Obviously there are degrees to rolling stops. one so slow that the driver can react easily (and is scanning so they can see the thing they need to react to) is fine, but some of the "rolling stops" ive seen in residential neighborhoods are crazy. Those definitely need to be made an example of.
Obviously thats when police discretion comes in. The police officer is the one issuing the ticket at the end of the day, so you need to trust that law enforcement wont be corrupt and pedantic. No amount of technology is gonna fix that
Yes, and to clarify that is the "rolling stop" I am talking about. Slow down, enough to verify that everything is clear, this will often mean coming to a nearly complete stop, especially if there are cars ahead of you or crossing. There's no need to come to a dead stop (and for how long? One second? Five? If you wait too long the driver on the cross street will get impatient and go out of turn). If I roll through a stop sign at a walking pace or slower that's not materially more unsafe than coming to a dead stop, and perhaps it is safer as it doesn't frustrate other drivers).
Of course if there are pedestrians waiting to cross, or the sight lines are bad, you behave accordingly.
Laws, rules, morays, norms etc. are in place for the "lowest common denominator", wether that be malicious people, people with impairments, older drivers, newer drivers etc. etc.
you as a human of course know not to hit a person walking thru an intersection. But a drunk person might think "eh I never fully stop and I don't see anyone".
We must all follow the rules to a TEE, ie; stopping even if completely clear, to signal to the lowest common denominators "this is the rule, you must stop regardless."
If this where not the case, by your logic, you can blow thru red lights, make left turns on red, drive against traffic etc. "as long as it's clear."
I personally am okay with enshittification of AI traffic cams if it promotes more aggressive traffic compliance. The police sure aren't.
If they actually cared about safety they'd license it to auto manufacturers and let people roll stop signs when it's safe and warm them when it's not. Or just put cheap traffic lights everywhere to speed up traffic. This is about earning revenue for municipallitues with micro enforcement zones.
(note/edit) I'm talking about flashing lights in the cab like when my car thinks I need to break. Forcing me to break unless I'm about to kill someone. Or just re-thinking the stop sign. The point of stop signs is they're effing cheap. If you're going to put AI cameras on all of them that is no longer cheap, could you not just turn them into lighted intersections that give the green to the right person and remove confusion and detect or have slappers for the pedestrians and just smooth out traffic everywhere? Or is the unsaid thing that stop signs are actually smoother because well you can roll them using your human brain to make decisions?
I feel like private companies enforcing when you can and cant move at a stop sign is a libertarian hellscape.
This isnt alot better, but at least its a provate vendor that gets data to the government who then decide to cite whats supposed to be dangerous behavior. Theres obviously corruption there, but these people are at least somewhat beholden tp the public through local elections and stuff. The toyota executives are not
Private companies enforcing the law with a profit motive is always a recipe for awfulness. Literally every other solution (including just not enforcing) is going to be a better one.
Jokes on them, my city doesn’t enforce cars without license plates very commonly visible [1]. So these plate readers are useless and people are regularly getting murdered on the streets with little to no consequences and to hit and run is the most advantageous position.
While I don’t doubt that OPD is ignoring plateless cars, that thread has outdated info. CA now does require temp plates, most places seem to be pretty good about enforcing the rule, and seeing a plateless car now seems (to me) to be a pretty good indicator that the person driving it is up to no good. My city further East in the East Bay is not exactly great about enforcing most traffic laws, but you’ll rarely see a car without a plate or temp these days.
Wondering what the opinion is about "Vision Zero." My little town is all over this. I think it's a bad idea. Goals should be realistic, and I think it's unrealistic to get to literally zero traffic deaths. There will always be random events leading to accidents and you can spend as much money as you want but you will never be able to prevent them all. At some point you're committing statistical murder by spending money that could be better used on ther things.
Vision Zero is based on a simple principle: if cars are driving less than 20mph a pedestrian collision is highly unlikely to be fatal.
So they set a speed limit of 20mph on any mixed use street, and create separated pedestrian/cyclist infrastructure for any street with higher speed limits.
The latter is super expensive, but it's what you need if you want usually zero fatalities and to go faster than 20mph.
Actually zero all the time is impossible, of course. It's possible that a 5mph collision with a frail pedestrian will kill them. So Norway and Sweden sometimes have a fatality. But a goal of "zero pedestrian fatalities most years" is actually feasible for polities with fewer than 10 million citizens.
No, certainly there is low hanging fruit. Just that "Vision Zero" isn't a realistic goal. I would have chosen a different name and goals to avoid pedantic debate on whether it's achievable.
> But instead of automating the entire setup, local governments review potential infractions before any citations are issued, ensuring a human is always in the loop.
IIRC, California abandoned automated traffic enforcement systems like these in the past because at the end of the day they were revenue negative. Having a human in the loop reduces the false positive rate, but drives up operating expenses to the point that it isn't sustainable.
Public services arent supposed to generate a profit, they are supposed to serve the public. If the system prevents people from being run over then its well worth the money
I think there's also the problem that the non-fiscal outcomes of red light cameras are mixed. There's this DOT research showing that red light cameras likely decreased right-angle collisions, but increased rear end collisions[1]. Studies weren't conclusive, but they seem to point to red light cameras not being a slam dunk that makes their deployment a no-brainer.
People don't want automated ticketing, so governments don't implement it.
In addition, there are many laws that aren't enforced and would generate instant outcry if they were. For example: it's illegal for someone of any age to ride a bicycle on the sidewalk, as opposed to the roadway, in many cities (in some cities it's the opposite).
If someone takes someone else's property without anyone seeing the act except a video camera, the video may be admitted as evidence.
Moving something into evidence that isn't testimony from a living breathing human is commonplace. A lawyer has someone attest to the authenticity and/or provenance of the item.
Another example of draconian enforcing the letter of the law in the name of making things "safer", then Goodhart's lawwing themselves into thinking they're succeeding. At least for myself, when I've got to deal with some kind of traffic control device and optimize my driving around that (say a speed bump), it takes my view/attention/focus away from looking for pedestrians elsewhere.
There's also the general problem with stop signs that if you do stop before the line as you're technically supposed to, then most times you can't actually see oncoming vehicle traffic. So most people stop over the line where they will be able to see, which means they're not planning on stopping where pedestrians walk. But fixing intersections is expensive meticulous work, while fining drivers for dealing with what they've been given is profitable.
If this were targeted at flagrant violations with warnings for a percentile of marginal cases (ie getting people who don't stop at all, and warning those who strain the idea of a rolling "stop"), then I could see it. But as it's worded, and as speed/red light cameras have been implemented, it just seems like another dynamic of a dystopian hellhole.
> There's also the general problem with stop signs that if you do stop before the line as you're technically supposed to, then most times you can't actually see oncoming vehicle traffic.
In these cases there's usually some other violation that is occurring, e.g., cars parked too close to the intersection, or shrubbery not properly cut back. As you note, the result is often they have to go beyond the stop sign. Even worse, I often end up so far in the street that if there was a car that was within its lane, but on the right-most side of it, they'd hit me (fortunately most drivers are aware enough to stop).
Sure, or the "violation" is by the city who would have to spend money to fix the intersection, thus they're incentivized to ignore it.
I think the general problem is that the presence of the stop sign and the 80-99% case (depending on the area/intersection) being to only worry about cars creates a type of blindness to pedestrians when they are there. Rather than these devices, I would say it would be more effective to install flashing lights on or around the stop sign, that turn on to signal a pedestrian is crossing.
(of course that would now be a bit of an uphill battle owing to the signs with the lights that flash all the time)
I recognized that in my last parenthetical sentence, as we've now got stop signs that flash all the time for similar "must do something" reasons. My point was specifically to only draw attention when it's warranted, to avoid that kind of fatigue. Like I think the crosswalks with flashing yellow lights when pedestrians are crossing are ridiculous overkill, but they don't add to my oversignage fatigue because they only demand that attention when the attention is needed.
The issue you’re highlighting is called the sight triangle. There are supposed to be restrictions on maximum landscaping height, etc, so you can see from the stop bar. Unfortunately many smaller crossings violate this.
In southern CA, traffic cameras were rolled out in tons of cities at basically every major intersection. They were a huge headache, did effectively nothing but waste taxdollars, and were scrapped for the purpose of issuing tickets. Except they weren't actually scrapped. They now feed data into several location-for-sale data brokers' pool which is queried by police. You're a little bit of a fool if you think this is about about "safety." Imagine the current license plate scanner tech combined with advanced facial recognition - if this isn't happening somewhere already - and tell me with a straight face cops aren't more excited about that dystopian future than stopping a few fender benders and generating meager city revenue (which they won't see anyway).
The dead giveaway to all these blatantly dishonest "safety" measures is they always, nearly without fail invoke safety "for the children." After all, who could be against that?
They don't even pretend it's about safety anymore. Flock cameras are all over the place and were never installed on the pretense of issuing citations or enforcing safety. It's all about tracking who is coming and going.
Itd be great if this sort of system could be trustworthy. How could that be done. Public data? But then that opens the data stream up to criminals who could stalk people and stuff. Third party audits? Who do you trust to do that? NGOs?
Remove the municipal revenue motive: no fines, only points on the license, with an option to plead not guilty in traffic court like a normal police interaction.
Points on the license probably requires more surveillance in order to prove who was driving rather than just that this particular vehicle violated a traffic law.
The whole idea that the way to reduce crime is by surveillance and enforcement is a con. Like in this case, all the places that managed to significantly reduce traffic accidents do so by carefully redesigning their street network to make safety easier and more intuitive.
Incentive structures need to be alligned better. Its so disheartening to see tech that works fine in europe fail in the US because of corruption and negative motivations.
What incentive structure could make these things be more benficial than just money grabbing? Laws that revenue from citations must be spend for direct public benefit only? With public audit?
As an aside, the U.S got roughly twice the number of fatal crash than the E.U [1][2], despite the E.U having ~100 millions more people.
There is a clear need to change a lot of things, whether it be (automatic) enforcement, redesigning infrastructure, and driver mentality.
[1] https://www.iihs.org/research-areas/fatality-statistics/deta...
[2] https://transport.ec.europa.eu/news-events/news/eu-road-fata...
I'm just speculating. But people in US in general drive a lot further for everything.
Most of shops i visit , other family members, my job , parks etc are all within about 5km of me.
The longest drive i do to visit a family member is 90km, every few months.
Anytime i've visited the US even going to the nearest shop seemed like a very long drive.
So if all other things are equal, theres just more opportunity for accidents.
Yes, this is why any statistics about driving that are meant to be compared across cultures need to have "distance driven" somewhere in the denominator.
Yea, I don't know I disagree. The denominator should always be just be population. I don't want to use sprawl as some sort of justification for more deaths per capita. Sprawl should be reduced and/or mitigated somehow.
I agree with the parent that distance driven should be in the equation. If there's a certain risk percentage associated with driving as a whole, then it's almost certainly somehow bound to distance driven.
For example, the risk of a car crash should be near zero if the distance driven is also zero. But even in the safest of vehicles, driving 100KM, 1,000KM, or 10,000KM+ each are going to have higher rates of accidents assuming real-life road conditions.
In the US, on average, people need to travel further to get to their destinations, and so on average I'd expect car accident and death statistics to likewise be higher.
I'm very in favor of reducing urban sprawl, utilizing public transit, etc., but in present state the US has vast quantities of existing road networks which can't be consolidated or mitigated overnight, so the next best way to improve those statistics would be improving traffic controls, vehicles, safety features, etc.
Measuring the wrong thing results in the wrong outcome.
Many of the most effective and efficient measures to reduce traffic fatalities work by reducing traffic. For example, promoting public transit. However, if you measure fatalities per kilometer these highly effective mechanisms have no effect on the metric.
It depends on which level of government we're talking about. The city governments can only really improve traffic controls.
But state and federal governments should absolute zero-in on reducing sprawl and travel distances and those actions should've started yesterday and they should stay laser focussed on reducing driving distances (especially for non-leisure reasons)
Different metrics measure different things. The right metric depends on what you want to compare and what kinds of claims you want to make, which determines what should be adjusted for in the metric.
So the obvious solution is driving less. Redesign the way of life to have to drive less.
"The longest drive i do to visit a family member is 90km, every few months."
In the USA, I have a friend who will do a tour visiting family every few months that takes him 1400 km outbound.
That's about 3 times the length of my country, i'd be in the sea lol.
EU has much higher training and licensing requirements in general.
In the US a 15 year old can get a learner's permit and start driving (with an adult) the same day. They can be licensed to drive on their own at 16 by passing a fairly cursory written exam and a short road test. No formal/classroom instruction is required.
Raising requirements in the US is prohibitive due to the lack of options individuals have when they cannot drive. A 16 year old often needs, or wants to work for an income. Owning a car is often a job requirement due to the distances involved in american cities. Getting a job within walking distance is often a non-option due to availability of employers, and biking is of limited safety profile depending on the area. If the only way to get to your job is through a state highway - driving may be the only safe method.
Roughly the same in Canada. I wish we had more stringent testing, including periodic re-testing (especially over 60).
For Germany there is even a subreddit about old people crashing into things: https://www.reddit.com/r/RentnerfahreninDinge/
Sadly it will be next to impossible to implement re-testing at a certain age as old people are the majority of the voters.
Lisence requirements vary by state. Here in Maryland, you need 30 classroom hours and 60 hours behind the wheel before before getting a provisional license.
The cost of getting a drivers license has increased a lot in Germany. 2500-3500€ isn't uncommon today.
Here in the NL I'd say it's at least €3600 if you have zero experience. This is my estimation for both theory & practical parts based on my own experience, current rates, and what little statistics I could find. Often much more if you fail and have to take more lessons.
Like everything in the US, those rules vary by state, and this is incorrect for at least those that I'm familiar with.
This is speculation on my part: I assume US citizens tend to drive more short distances within cities to get to places, like shopping, while many EU cities are walk-able.
Cities and city entrances have the largest concentration of people and accidents.
I wonder if the size of the cars matter. Two Honda Fits crashing into each other is going to be different than a Honda Fit and a Ford 150.
I can't speak to every European country, but Portugal also has a lot fewer stop signs and traffic lights than the US (roundabouts are one reason, but there are multiple four-way intersections on my street that would have stop signs in the US).
Given the way American streets are set up, rigorously enforcing stop signs is probably beneficial, but I think other factors about how streets are arranged and how people drive are more important.
Possible causes for this include the prevalence of really large SUVs, which make it physically much more difficult to even see pedestrians - especially children.
Another part is truck design. Same reason: American trucks have elongated noses for the engine, whereas European trucks have the driver sitting directly above the engine.
On top of that, European countries have much more strict testing requirements on vehicles. Basically, every 2-4 years you have to have your vehicle inspected for roadworthiness - foundational stuff such as structural rust, worn-down tires or brakes gets caught much, much earlier than in the US.
To be fair to the US, various states have yearly inspections that include undercarriage rusting and other issues. It's a factor but I think less important than the truck part.
Fatalities are horrible to the people involved, and in many cases the people responsible for such are punished. This is about saving a fractional percent of people while backdooring the technologies as a surrogate for control of everyone else.
The error in agreeing to automatic enforcement lay in the indirect failures that naturally follow within centrally structured systems, when those automatic systems stop working correctly, or worse selectively work; the world will be worse off than not having the solution in place at all.
There are dramatically more risks of this becoming a component of a panopticon prison in a fascist state, something the US is degrading into right now with the slow erosion; and stress fractures to our rule of law, it might very well suddenly fail overnight.
What impact will these solutions have in breeding discontent if everyone has the boot of the government on their neck every time they roll through an empty stop-sign where no one is there..., or worse when they did stop and the AI mis-categorized it as a rolling stop. What feedback systems correct a faulty running system? Government and government apparatus have trouble getting sufficient benefits to legitimate welfare recipients, what makes you think they'll do this any better? Competency is not a common trait for government workers.
Who do you think will be most impacted, the people with less awareness, or the people with more awareness. Lower IQ/cognitive speed vs. Higher IQ/cognitive speed. Would this result in an evolutionary filter against intelligence?
Would these dramatic changes drive the intelligent people which society rests upon (dependently so), so crazy that they end themselves, don't have children, or end their children and themselves? Is there any hope for a future under such repressive and stagnant systems. No there isn't. Intelligent thought is largely based in cognitive speed, and multiplied by education. There are some very educated people who are not necessarily sufficiently intelligent to stand in for these people. Their words and ideas often cause more harm, the more complex the system becomes.
The moment you rest an argument on do it for the dead people, or do it for the children, which is what %, you dismiss all the failures of the proposed system. Those failures still occur, those harms still happen, and the type of people you have left are less capable of adapting, or rather become enraged with each additional reminder that they are not people, they are slaves or animals.
A nation becomes strong only as a result of its strong people in unity.
When you make people necessarily dependent on the imagined detriment of what could happen, prevent them from acting, and do this at the expense of what is actually happening, you get a weak fragile complacent brittle people who break and are parasitically dependent on a pool of people that shrinks to nothing.
These detrimental characteristics spread over time both laterally among people but also generationally, and eventually circumstances occur where your people simply cannot adapt to what the environment requires as needed, and in that existential threat you face oblivion as a species, extinction.
Complacency, and a blindness or reactance to the risks, breed delusion which takes root spreading to those that remain, as a contagion.
The moment you think you can make people better by treating them like animals or slaves, or prisoners, is the inflection point towards your people's ultimate destruction; although it may be many years between. Every person is dependent on every other person indirectly, and some carry more than others.
How do you suppose such camera's of an all seeing eye will change the populace for the worse, might it make them more animalistic, ugly, violent... just as Sauron did as described by Tolkien, and much of the basis for Tolkien's works is based on the bible.
The only way to win a game of thermonuclear war is to not play the game. The same can be said about a lot of decisions which pigeonhole your future into a box without a future.
Police no longer feel the need to do their jobs, and Americans in general have just lost any sense of empathy or even awareness of other people.
But also we have a serious problem where taking away someone's license to drive is to sentence them to poverty if not homelessness and starvation. We don't have decent public transit and there are very few jobs within walking distance of most residential areas. Those jobs that do exist don't pay a living wage because pegging minimum wage to inflation or even the poverty line is "communism" and an "attack on businesses".
Our problems with car fatalities is really only one small symptom of the ongoing collapse of American society.
My objection to raising minimum wage too high is that it makes it illegal for people with modest skills and abilities to earn a living unless they can find an employer to dress up charity as a job.
I know people who would struggle to create $100 of value for an employer per day. I would rather they be allowed to hold a job at $80/day than to have minimum wage set above the value they are able to create.
Nothing wrong with a "rolling stop" if the sight lines are good and there are no pedestrians or crossing traffic. The point of a stop is to allow traffic to cross the intersection in a safe and orderly fashion. If you slow down, and verify that everything is clear, then that objective is achieved even if you don't come to a complete stop.
If these cameras were smart enough to issue citations when pedestrians or cross-traffic is present I could support it. But issuing a citation at a deserted intersection when no risk is created is just absurd.
Then cities should adopt the yield signs that say that. I agree our system could be much smarter. So may timer-driven systems would be better if the computer knew the presence and number of cars, pedestrians, bikes, etc.
I could support this if you combined it with criminal and civil liability when you guess wrong and run someone over while blowing your stop-sign. Right now, that's a $500 ticket at best, and it happens every day.
The whole problem is that people don't look for pedestrians -- they look for another car that might hit them. So they are looking the wrong way. And then they tell the cops some sob-story about how the dead pedestrian "came out of nowhere".
Um, no. If you hit and injure someone in a crosswalk you can be sure you will face civil liability. Hope you have enough insurance (the legally required minimums are far too low).
> If you slow down, and verify that everything is clear, then that objective is achieved even if you don't come to a complete stop.
There are too many failure points there to trust mediocre meat sacks to follow that process correctly. Remember that driving rules and restrictions are not written assuming an alert, effective, and skilled driver operating a well maintained vehicle, they are written assuming an average person who has successfully completed a driver's test driving something that passes basic road worthiness checks.
It works well enough in other countries than America.
E.g. in the Netherlands or Germany there's no need for 4-way stops and such. If no other signage applies, then whoever is on the "right" side has priority over the people on the "left" side. And exceptions do apply, i.e. it's not "that simple" either. It does depend on whether both roads are on the same level or not. A road to your right that has a sidewalk border stone running across it does not give them the right of way, while if the sidewalk border stops and both roads intersect directly at the same level, then the road to your right does have right of way.
So e.g. if you take a typical urban development with lots of little streets and houses, where you'd see a lot of rolling stops in America, nobody's gonna stop at every intersection, rolling or not there. This does go as far as when cars from all directions arrive at the same time, then nobody has automatic right of way and one of them has to wave the person to their right through and will be the last allowed to proceed.
> Nothing wrong with a "rolling stop" if the sight lines are good and there are no pedestrians or crossing traffic.
Why bother rolling the stop, it should be ok to blow through it at full speed if you're sure it's clear.
It's about the incremental injury / death; full speed is going to cause more. Going through at 3mph shouldn't cause more, but if rolling statistically does then full stop should be enforced.
But what is the point of rolling through at 3mph? To save 2 seconds? The reality is people roll stops because they are a bit lazy, and stopping fully requires marginally more effort. And while that laziness is harmless at empty intersections, it invariably turns into complacency and habit that bleeds into situations that aren't as safe. The same dynamic happens with signaling lane changes.
This works for emergency vehicles
The problem is that you can't always tell that there are no pedestrians or crossing traffic unless you take enough time to come to a complete stop. There are plenty of stop signs where that's not true, but also plenty where it is, and it's not always clear which one it is to the driver.
I think the right fix here isn't total enforcement nor relaxed enforcement, but relaxed signage. If sight lines are good enough that it's safe to roll through, that should be a yield, not a stop. Stop should mean, you actually need to come to a complete stop to safely navigate this intersection. Then you can enforce it without qualms.
Good point, another thing that the EU tends to do differently from the US. US rarely uses "Yield." The only places I generally see it are at roundabout entries.
Often on EU roads they will use "sharks teeth" yeild markers where a side road enters or crosses a main road. The requirement there is essentially "proceed if clear" a full stop is not required. I have rarely (maybe never) seen a US-style 4-way stop there (my experience is limited to Scandinavia and Germany).
In the UK, driving a decade all over, never seen a 4 way stop.
Signage is ineffective in addressing short term environmental or visibility impacts. Sure, it might be easy to see during the day with clear visibility. What about at night? Fog? Snow or rainstorm that is restricting visibility? Some dropped a storage pod on the road that obstructs the view of everything except a yield sign?
> If sight lines are good enough that it's safe to roll through, that should be a yield, not a stop
They are
Definitely not, there are tons of stop signs in places with low speeds and great sight lines that are perfectly safe to treat as a yield. And conversely, there are occasionally some interesting places (on-ramps for very old freeways) that have a yield where it's not safe to proceed without stopping first.
I would bet on traffic planners’ assessment of these variables to be more reliable and comprehensive than yours, in general.
Spoken like someone who has never driven a car.
Nah, spoken like someone repeatedly humbled by the complexity and detail of domains other than my own.
“I can drive a car therefore I understand traffic design better than traffic designers” is obviously an absurd statement when you just say it outright instead of condescendingly implying it.
> HN commentator revolutionizes traffic design with groundbreaking insight: consider the sight lines
I doubt anyone is even analyzing most of these intersections. Low-traffic residential streets just automatically get a stop sign regardless.
With computer vision the case of checking for pedestrians in the vicinity is trivial. So these cameras are definitely worth it for that
I do disagree about the rolling stop though. After drunk driving, drivers getting too relaxed and working off of predictive execution has to be the biggest cause of road accidents. A driver rolling past a stop at high speed in a school zone cant react fast enough to kids running past or even just walking on predictive execution themselves because they think the car will stop.
Obviously there are degrees to rolling stops. one so slow that the driver can react easily (and is scanning so they can see the thing they need to react to) is fine, but some of the "rolling stops" ive seen in residential neighborhoods are crazy. Those definitely need to be made an example of.
Obviously thats when police discretion comes in. The police officer is the one issuing the ticket at the end of the day, so you need to trust that law enforcement wont be corrupt and pedantic. No amount of technology is gonna fix that
Yes, and to clarify that is the "rolling stop" I am talking about. Slow down, enough to verify that everything is clear, this will often mean coming to a nearly complete stop, especially if there are cars ahead of you or crossing. There's no need to come to a dead stop (and for how long? One second? Five? If you wait too long the driver on the cross street will get impatient and go out of turn). If I roll through a stop sign at a walking pace or slower that's not materially more unsafe than coming to a dead stop, and perhaps it is safer as it doesn't frustrate other drivers).
Of course if there are pedestrians waiting to cross, or the sight lines are bad, you behave accordingly.
Laws, rules, morays, norms etc. are in place for the "lowest common denominator", wether that be malicious people, people with impairments, older drivers, newer drivers etc. etc.
you as a human of course know not to hit a person walking thru an intersection. But a drunk person might think "eh I never fully stop and I don't see anyone".
We must all follow the rules to a TEE, ie; stopping even if completely clear, to signal to the lowest common denominators "this is the rule, you must stop regardless."
If this where not the case, by your logic, you can blow thru red lights, make left turns on red, drive against traffic etc. "as long as it's clear."
I personally am okay with enshittification of AI traffic cams if it promotes more aggressive traffic compliance. The police sure aren't.
Spoken like a C programmer!
It's just optimizing the reward function. The traffic ticket maximizer is the paperclip maximizer's sibling.
If they actually cared about safety they'd license it to auto manufacturers and let people roll stop signs when it's safe and warm them when it's not. Or just put cheap traffic lights everywhere to speed up traffic. This is about earning revenue for municipallitues with micro enforcement zones.
(note/edit) I'm talking about flashing lights in the cab like when my car thinks I need to break. Forcing me to break unless I'm about to kill someone. Or just re-thinking the stop sign. The point of stop signs is they're effing cheap. If you're going to put AI cameras on all of them that is no longer cheap, could you not just turn them into lighted intersections that give the green to the right person and remove confusion and detect or have slappers for the pedestrians and just smooth out traffic everywhere? Or is the unsaid thing that stop signs are actually smoother because well you can roll them using your human brain to make decisions?
I feel like private companies enforcing when you can and cant move at a stop sign is a libertarian hellscape.
This isnt alot better, but at least its a provate vendor that gets data to the government who then decide to cite whats supposed to be dangerous behavior. Theres obviously corruption there, but these people are at least somewhat beholden tp the public through local elections and stuff. The toyota executives are not
Private companies enforcing the law with a profit motive is always a recipe for awfulness. Literally every other solution (including just not enforcing) is going to be a better one.
Jokes on them, my city doesn’t enforce cars without license plates very commonly visible [1]. So these plate readers are useless and people are regularly getting murdered on the streets with little to no consequences and to hit and run is the most advantageous position.
[1] https://www.reddit.com/r/oakland/comments/4wdd57/whats_the_d...
While I don’t doubt that OPD is ignoring plateless cars, that thread has outdated info. CA now does require temp plates, most places seem to be pretty good about enforcing the rule, and seeing a plateless car now seems (to me) to be a pretty good indicator that the person driving it is up to no good. My city further East in the East Bay is not exactly great about enforcing most traffic laws, but you’ll rarely see a car without a plate or temp these days.
If your idea of solving a problem is "how can we use technology to control people's behavior", you are a net negative value to humanity, full stop.
Self driving cars are the real solution to this problem and the only solution to this problem. And maybe also teach your kids to look both ways.
Wondering what the opinion is about "Vision Zero." My little town is all over this. I think it's a bad idea. Goals should be realistic, and I think it's unrealistic to get to literally zero traffic deaths. There will always be random events leading to accidents and you can spend as much money as you want but you will never be able to prevent them all. At some point you're committing statistical murder by spending money that could be better used on ther things.
Vision Zero is based on a simple principle: if cars are driving less than 20mph a pedestrian collision is highly unlikely to be fatal.
So they set a speed limit of 20mph on any mixed use street, and create separated pedestrian/cyclist infrastructure for any street with higher speed limits.
The latter is super expensive, but it's what you need if you want usually zero fatalities and to go faster than 20mph.
Actually zero all the time is impossible, of course. It's possible that a 5mph collision with a frail pedestrian will kill them. So Norway and Sweden sometimes have a fatality. But a goal of "zero pedestrian fatalities most years" is actually feasible for polities with fewer than 10 million citizens.
There's a verse I like a lot in Tao Te Jing:
"One must know when it is enough. Those who know when it is enough will not perish."
I think it's good to aim for zero traffic deaths, as in many countries the situation is so bad that a lot of improvement is feasible.
The long tail would definitely be much harder to tackle, but I don't expect this to be a serious problem in practice.
This post is about ticketing people who run stop signs. Is that the kind of price you consider too high?
No, certainly there is low hanging fruit. Just that "Vision Zero" isn't a realistic goal. I would have chosen a different name and goals to avoid pedantic debate on whether it's achievable.
> But instead of automating the entire setup, local governments review potential infractions before any citations are issued, ensuring a human is always in the loop.
IIRC, California abandoned automated traffic enforcement systems like these in the past because at the end of the day they were revenue negative. Having a human in the loop reduces the false positive rate, but drives up operating expenses to the point that it isn't sustainable.
Public services arent supposed to generate a profit, they are supposed to serve the public. If the system prevents people from being run over then its well worth the money
Oddly enough, policing as a business leads to some undesirable outcomes.
I think there's also the problem that the non-fiscal outcomes of red light cameras are mixed. There's this DOT research showing that red light cameras likely decreased right-angle collisions, but increased rear end collisions[1]. Studies weren't conclusive, but they seem to point to red light cameras not being a slam dunk that makes their deployment a no-brainer.
[1] https://www.fhwa.dot.gov/publications/research/safety/05049/
Technology is not the roadblock here.
People don't want automated ticketing, so governments don't implement it.
In addition, there are many laws that aren't enforced and would generate instant outcry if they were. For example: it's illegal for someone of any age to ride a bicycle on the sidewalk, as opposed to the roadway, in many cities (in some cities it's the opposite).
It is generally illegal in the USA to be accused of a crime without being "confronted by the witnesses against him".
Red light cameras foundered on that obligation since they were generally run out of State and fly the camera operators in was not cost effective.
Also, just because your car broke the law doesn't mean you did, which was another defense that worked.
I'm surprised that the citations aren't thrown out...
If someone takes someone else's property without anyone seeing the act except a video camera, the video may be admitted as evidence.
Moving something into evidence that isn't testimony from a living breathing human is commonplace. A lawyer has someone attest to the authenticity and/or provenance of the item.
> “Ultimately, we hope our technology becomes obsolete,” says Maheshwari.
They may, but once something becomes a revenue stream, their successors won't.
Another example of draconian enforcing the letter of the law in the name of making things "safer", then Goodhart's lawwing themselves into thinking they're succeeding. At least for myself, when I've got to deal with some kind of traffic control device and optimize my driving around that (say a speed bump), it takes my view/attention/focus away from looking for pedestrians elsewhere.
There's also the general problem with stop signs that if you do stop before the line as you're technically supposed to, then most times you can't actually see oncoming vehicle traffic. So most people stop over the line where they will be able to see, which means they're not planning on stopping where pedestrians walk. But fixing intersections is expensive meticulous work, while fining drivers for dealing with what they've been given is profitable.
If this were targeted at flagrant violations with warnings for a percentile of marginal cases (ie getting people who don't stop at all, and warning those who strain the idea of a rolling "stop"), then I could see it. But as it's worded, and as speed/red light cameras have been implemented, it just seems like another dynamic of a dystopian hellhole.
> There's also the general problem with stop signs that if you do stop before the line as you're technically supposed to, then most times you can't actually see oncoming vehicle traffic.
In these cases there's usually some other violation that is occurring, e.g., cars parked too close to the intersection, or shrubbery not properly cut back. As you note, the result is often they have to go beyond the stop sign. Even worse, I often end up so far in the street that if there was a car that was within its lane, but on the right-most side of it, they'd hit me (fortunately most drivers are aware enough to stop).
Sure, or the "violation" is by the city who would have to spend money to fix the intersection, thus they're incentivized to ignore it.
I think the general problem is that the presence of the stop sign and the 80-99% case (depending on the area/intersection) being to only worry about cars creates a type of blindness to pedestrians when they are there. Rather than these devices, I would say it would be more effective to install flashing lights on or around the stop sign, that turn on to signal a pedestrian is crossing.
(of course that would now be a bit of an uphill battle owing to the signs with the lights that flash all the time)
Yea, then we have flashing lights all over the place and people become blind to them, too.
I recognized that in my last parenthetical sentence, as we've now got stop signs that flash all the time for similar "must do something" reasons. My point was specifically to only draw attention when it's warranted, to avoid that kind of fatigue. Like I think the crosswalks with flashing yellow lights when pedestrians are crossing are ridiculous overkill, but they don't add to my oversignage fatigue because they only demand that attention when the attention is needed.
The issue you’re highlighting is called the sight triangle. There are supposed to be restrictions on maximum landscaping height, etc, so you can see from the stop bar. Unfortunately many smaller crossings violate this.
For example see fig. 71 at https://highways.dot.gov/safety/other/older-road-user/handbo...
In southern CA, traffic cameras were rolled out in tons of cities at basically every major intersection. They were a huge headache, did effectively nothing but waste taxdollars, and were scrapped for the purpose of issuing tickets. Except they weren't actually scrapped. They now feed data into several location-for-sale data brokers' pool which is queried by police. You're a little bit of a fool if you think this is about about "safety." Imagine the current license plate scanner tech combined with advanced facial recognition - if this isn't happening somewhere already - and tell me with a straight face cops aren't more excited about that dystopian future than stopping a few fender benders and generating meager city revenue (which they won't see anyway).
The dead giveaway to all these blatantly dishonest "safety" measures is they always, nearly without fail invoke safety "for the children." After all, who could be against that?
They don't even pretend it's about safety anymore. Flock cameras are all over the place and were never installed on the pretense of issuing citations or enforcing safety. It's all about tracking who is coming and going.
Itd be great if this sort of system could be trustworthy. How could that be done. Public data? But then that opens the data stream up to criminals who could stalk people and stuff. Third party audits? Who do you trust to do that? NGOs?
Remove the municipal revenue motive: no fines, only points on the license, with an option to plead not guilty in traffic court like a normal police interaction.
Points on the license probably requires more surveillance in order to prove who was driving rather than just that this particular vehicle violated a traffic law.
The whole idea that the way to reduce crime is by surveillance and enforcement is a con. Like in this case, all the places that managed to significantly reduce traffic accidents do so by carefully redesigning their street network to make safety easier and more intuitive.
whenever a cop car is around, everyone becomes grandma and starts driving 5 under. very annoying.
didn't get past the "Unblinking eyes could lower the vehicular death toll". The unblinking eye is there to maximize citations, end of story.
Incentive structures need to be alligned better. Its so disheartening to see tech that works fine in europe fail in the US because of corruption and negative motivations.
What incentive structure could make these things be more benficial than just money grabbing? Laws that revenue from citations must be spend for direct public benefit only? With public audit?