While I'm all for reducing pedestrian deaths, this seems underwhelming. Based on their own numbers, this could save 67 lives a year. In 2022, there were 7522 deaths, up from a low in 2009 of 4109...
While I'm all for reducing pedestrian deaths, this seems underwhelming.
Based on their own numbers, this could save 67 lives a year. In 2022, there were 7522 deaths, up from a low in 2009 of 4109 deaths (source). Is there no other change that they could propose that would reduce fatalities by more than 1%? Given the ever increasing number of pedestrian deaths each year, will a 1% change even show up in the data?
Pedestrian deaths are largely due to a combination of infrastructure, distracted driving, and increased vehicle sizes. The NHTSA has the power to force the industry to change significantly. Make vehicles smaller. Make vehicles spy on us to make sure we're not looking at your phone. Do something.
If this were anything other than the auto industry, people would be angry about it. 7522 deaths per year is a lot of bodies. And that pile is only getting larger.
There's no reason this can't be done fully locally in the car. Driving a multi-ton machine that can easily kill others means that you have restrictions placed upon you. This has always been part...
There's no reason this can't be done fully locally in the car.
Driving a multi-ton machine that can easily kill others means that you have restrictions placed upon you. This has always been part of the contract you make with the state to be allowed to drive a car. For example, you can be required to have blood drawn for testing if you are suspected of drunk driving, something that is infinitely more invasive.
But infinitely more infrequent and requiring probable cause, and only rarely done without consent. I'd not trust the car to be able to actually see where I'm looking, in sunglasses, in my regular...
But infinitely more infrequent and requiring probable cause, and only rarely done without consent.
I'd not trust the car to be able to actually see where I'm looking, in sunglasses, in my regular glasses, or for folks with different skin tones or eye shapes, historically we're not great at this.
If we go this route, out a breathalyzer lock on every car too. No driving if you have any alcohol in your system.
I'm not arguing for distracted driving or drunk driving but I do think this is way less reasonable than demanding safer cars.
In theory, such a system could just check to see if you're holding a phone (or anything) in your hand, whether or not you're looking at it. That would cover about 95% of distracted drivers today...
I'd not trust the car to be able to actually see where I'm looking, in sunglasses, in my regular glasses, or for folks with different skin tones or eye shapes, historically we're not great at this.
In theory, such a system could just check to see if you're holding a phone (or anything) in your hand, whether or not you're looking at it. That would cover about 95% of distracted drivers today and it's simpler and less error-prone than eye-tracking.
If we go this route, out a breathalyzer lock on every car too. No driving if you have any alcohol in your system.
That's actually already in the works - mandatory in every new car starting in 2026.
Holding something in my hand is gonna have a lot of false positives too. And, much as I'm not sold on breathalyzer vehicles, they don't seem like they'd actually lock you out, it's just a check...
Holding something in my hand is gonna have a lot of false positives too.
And, much as I'm not sold on breathalyzer vehicles, they don't seem like they'd actually lock you out, it's just a check before you drive for your information. Hate it from a "clearly it'll be subpoenaed" standpoint. Maybe useful from a "I didn't think I was buzzed" standpoint. Not that useful from an "intentionally drunk driving" standpoint.
So what's the car gonna do, beep at me when I grab a sip of my drink, or put my hand on my thigh, or wave my arm around singing on a road trip? It'll get turned off so fast by anyone who can read a manual.
Once again, I almost never drink and am wary about driving after half a cider at a restaurant. I do work very hard not to get distracted by my phone, using audiobooks/podcasts and navigation on every trip to keep my brain busy and able to focus. I also haven't held my phone in my hand while driving for probably a decade now.
Anyway I'm not actually that pressed, because I don't think the tech is manageable, but I think it's the wrong focus for pedestrian safety contrasted with speed and vehicle size.
This will not happen. This is very rare, usually replaced with a breath test, arguably less invasive for many since it is one-time (rather than constant) and people do not often have to worry...
There's no reason this can't be done fully locally in the car.
This will not happen.
For example, you can be required to have blood drawn for testing if you are suspected of drunk driving, something that is infinitely more invasive.
This is
very rare,
usually replaced with a breath test,
arguably less invasive for many since it is one-time (rather than constant) and people do not often have to worry about the contents of their own blood,
Bad driving, including driving under the influence, can be easily prevented without having to resort to a surveillance state. Advocating for one to be implemented because you're afraid of other people's driving is insane.
There’s no reason an Surveillance state needs to be horrifying. It’s still not even close to worth the risk. Edit: Now that I have some more time, hell there's LOTS of reasons that doing it...
There's no reason this can't be done fully locally in the car.
There’s no reason an Surveillance state needs to be horrifying. It’s still not even close to worth the risk.
Edit:
Now that I have some more time, hell there's LOTS of reasons that doing it locally in the car is a fucking problem.
What exactly does the system do on a positive identification of a phone use in the car? Pull the vehicle over? Stop it? Auto send a rate increase to your insurance?
How in the HELL do you control for false positives? The image detection software i've seen is impressive. It's not, "risk my life or my future" impressive.
How many people are you willing to punish? Shall we include all distracted driving? Single mother who took her eyes off the road to give her kid a bottle? Parent rushing to their child while on the phone as they're having some sort of major incident?
How much money are you willing to force everyone to spend? Cars are more and more becoming computers on wheels, but this sort of real time detection system, fully local, is going to cost a boatload more than any current system.
How much are you willing to punish drivers who break or mess with the system? You can drive with all sorts of damage to your vehicle, arguably legally depending on what kind. Am I unable to drive my car in an emergency because my detection system is damaged or broken? Oops a lens cracked or got covered so now i'm unable to drive? Or do you allow it and now you've just got a bunch of false positives again or people who already can barely afford to hang on getting punished even worse?
There's all sorts of grey area's in this that you're just asking to fight with. These idea's are always half baked and dangerous before you even get to the "and the state will 100% abuse this" scenarios.
This is already solved by existing partial self-driving technologies: nudge the wheel a bit to test the driver's attention, and if they don't respond in time, slow the vehicle down and pull over....
What exactly does the system do on a positive identification of a phone use in the car? Pull the vehicle over? Stop it? Auto send a rate increase to your insurance?
This is already solved by existing partial self-driving technologies: nudge the wheel a bit to test the driver's attention, and if they don't respond in time, slow the vehicle down and pull over. As a bonus, this would save the lives of drivers who are having a medical emergency as well.
How in the HELL do you control for false positives? The image detection software i've seen is impressive. It's not, "risk my life or my future" impressive.
Test the driver's attention subtly.
How many people are you willing to punish? Shall we include all distracted driving? Single mother who took her eyes off the road to give her kid a bottle? Parent rushing to their child while on the phone as they're having some sort of major incident?
I think I've addressed this above but I want to note that driving is the only place where this cavalier attitude to safety exists. I may just have been lucky, but I've never been in any job site or warehouse where people operating heavy machinery gave it anything but their full attention. But cars are styled like toys, not heavy machinery, and driver licensing in this country is a complete joke, so people don't take them seriously like they should.
What you're describing here is known as a "normalization of deviance". This is a classic human fallacy. Our attitudes to driving would not pass muster anywhere that safety is taken seriously.
How much money are you willing to force everyone to spend?
Even if it didn't fit inside existing systems, I'd expect the overall cost of this kind of system to be < $50.
How much are you willing to punish drivers who break or mess with the system?
We already have this mostly figured out with our emissions systems. Something like 2-5% insist on bypassing regulations, but having the vast majority of people leaving things alone is a huge improvement.
So this system not only needs to identify phone use, but also if they're having a medical emergency? What happens if i'm on my phone and it nudges the wheel and I respond? I'm good to go? Can the...
So this system not only needs to identify phone use, but also if they're having a medical emergency?
What happens if i'm on my phone and it nudges the wheel and I respond? I'm good to go? Can the driver override the pull over, because again, false positives are now actively dangerous to the driver and the others on the road.
You are vastly overestimating the abilities of self driving systems and vastly underestimating the cost of them. You are talking about interior and exterior camera and detection devices along with the system required to handle and process. Further basically every self driving system out there can't handle ANY deviation, so construction, rain, snow, unknown objects, etc are all situations where this solution doesn't work in the best case, and is actively dangerous in the worst case.
You also seem to be ignoring that the reason that cars are treated differently from heavy machinery is because they are 100% necessary in some situations. Many urban and country areas require a car to function, and it's already a huge burden on the poor. Having a malfunctioning safety system on a piece of work equipment is a call to the repair tech and ideally you've got a backup. The same on a car, as you've proposed, is going to be absolutely devastating to people who already can't afford to keep their cars in perfect condition.
This is before you get into cases where the car refuses to drive and someone dies because of it.
That's a reasonable position, but there's a good chance that in the near future our cars are going to spy on us anyway. Might as well put that to good use.
That's a reasonable position, but there's a good chance that in the near future our cars are going to spy on us anyway. Might as well put that to good use.
No. That's how these things always happen. Every single advancement of mass surveillance and erosion of rights is excused by some appeal to how something bad can be lessened. And the plan is...
No. That's how these things always happen. Every single advancement of mass surveillance and erosion of rights is excused by some appeal to how something bad can be lessened. And the plan is always this nebulous.
It needs to be a criminal act to record people without reasonable consent, with a similar test to the GDPR: you can't make acceptance of the invasion of privacy a requirement to use a product that doesn't reasonably need that to function. And that goes doubly for government.
If you want to reduce pedestrian fatalities, one obvious thing has changed between 2013 and the present, in the period where pedestrian fatalities rose almost 60%: the average car will strike you in the upper chest and drive over you instead of hitting your knees and causing you to fall onto the hood. That's the thing that needs to be corrected.
If you want something concrete, in California Senate Bill 961 (To require passive speeding notifications in cars) is currently on the Governor's desk. If you're a resident, calling and asking him...
If you want something concrete, in California Senate Bill 961 (To require passive speeding notifications in cars) is currently on the Governor's desk. If you're a resident, calling and asking him to sign would be very helpful. If it does pass, because California is such a significant market it's likely to do a lot to shape the likelihood of nationwide adoption of similar measures in new cars.
I'm more and more in favor of automated ticketing on toll roads. I used to be very against it, feeling it too authoritarian, but I've come around because car ownership needs to be heavily...
I'm more and more in favor of automated ticketing on toll roads. I used to be very against it, feeling it too authoritarian, but I've come around because car ownership needs to be heavily regulated anyhow, because cars are large dangerous pieces of machinery.
For speeding on highways in particular, it's also much less likely to have context-dependent wiggle room, false positives, and unintended consequences than something like red-light cameras, simply because physics dictates the limits. You measure the time between point A and point B and calculate the fee based on how much over the speed limit you did on average. While this wouldn't eliminate speeders on the whole, it would certainly drive a large enough segment of drivers, professionals in particular, that the speeders would be forced to drive the speed limit courtesy of congestion.
Automatic speed cameras are another thing happening in the Bay Area, though it's not mostly on highways. They're part of a package of improvements to help reduce speeding around pedestrians, as...
Automatic speed cameras are another thing happening in the Bay Area, though it's not mostly on highways. They're part of a package of improvements to help reduce speeding around pedestrians, as most of the vehicle-related injuries and deaths aren't from highway traffic, but in places where people trying to travel in various ways intersect. Cars are at their safest when everyone is moving in parallel lines with minimal variation in capabilities.
Oh the highway thing is beyond just reducing deaths, but also emissions. It tends to bleed in whenever I think about these things...especially since the necessary infrastructure is already...
Oh the highway thing is beyond just reducing deaths, but also emissions. It tends to bleed in whenever I think about these things...especially since the necessary infrastructure is already completely in place on many highways.
Additionally, while it may be feasible in the Bay area, I can think of immense difficulty elsewhere having automated speeding tickets provide any tangible benefit without being overbearing to the point that people pass laws to ban them. You'd have to have checkpoints almost every intersection in a grid to be effective and that borders closer to 'tracking your every move from start to finish' than 'enforcing a baseline rule on major arteries.' If anything, having automated speeding in some places and not others off-highway could have adverse effects on passenger deaths as speeders try to figure out routes to avoid automatic ticketing. Why take the 45 mph road with automated speeding detection when you can blow through a 25mph residential area at 60 mph and be less likely to get caught?
However, I do think that killing speeding on highways will also reduce deaths off-highways, as conditioning people to obey the speed limit on highways will carry over to not-speeding everywhere.
While I have not dug in, I'd say that there are almost certainly few side benefits that will come along for the ride. Cars will almost certainly start trending downward in size....reducing the...
While I have not dug in, I'd say that there are almost certainly few side benefits that will come along for the ride. Cars will almost certainly start trending downward in size....reducing the weight of a vehicle will be far easier to come into compliance than most other methods. This has added benefits of reducing wear on roads and reducing non-pedestrian deaths as well. One other I'd like to see NHTSA is to have mandatory governors that put a max speed of 70mph on car-exclusive roads like highways, and a max speed of 45mph on any mixed-use road. Or perhaps a mandatory-stop function when crosswalks are active. I very much like that one....all crosswalks emit a powerful signal when active, and if the car detects that signal at a ~10ft range engages emergency braking. There would be a lot of fender benders to start, but driver behavior would adjust pretty rapidly with that one.
While this isn't directly addressable by NHTSA, states have the most power to force drivers to chose smaller vehicles. One of the best rules we could get to trend vehicle size downward rapidly is increasing licensing difficulty depending on fully-laden GVW. Split off consumer vehicles into more classes, with more stringent road testing to be permitted to drive larger vehicles. This would allow light-vehicle licenses to have longer expiration duration and 0-cost, while heavy-vehicle licenses need more regular road tests. Having this clear separation could also allow increasing fines for moving violations based on the highest-weight vehicle you are licensed to drive. Having a speeding ticket (or run stoplight) be $20 if you're only authorized up to 1t vehicle, but $5000 for a 3t vehicle would force demand for smaller vehicles quite quickly.
If actual infrastructure is being developed and installed to solve these problems, then they can just skip directly to self-driving cars. Better yet, use that money for improved public transit and...
Or perhaps a mandatory-stop function when crosswalks are active. I very much like that one....all crosswalks emit a powerful signal when active, and if the car detects that signal at a ~10ft range engages emergency braking.
If actual infrastructure is being developed and installed to solve these problems, then they can just skip directly to self-driving cars. Better yet, use that money for improved public transit and intercity HSR.
One of the best rules we could get to trend vehicle size downward rapidly is increasing licensing difficulty depending on fully-laden GVW. Split off consumer vehicles into more classes, with more stringent road testing to be permitted to drive larger vehicles
I do like this suggestion. It's kind of insane that I took my driving test in a subcompact car and now I can legally drive a motorhome towing that same car, or drive a gigantic pickup truck with a 5th wheel hookup and tow a 30 foot RV trailer.
The thing is though, it would be somewhat trivial to design such a signaling mechanism that just wires in directly to the crosswalk light, possibly even in-line with the bulb itself. Doesn't...
The thing is though, it would be somewhat trivial to design such a signaling mechanism that just wires in directly to the crosswalk light, possibly even in-line with the bulb itself. Doesn't really need to be more than a passive BLE beacon, which we have cellphones able to detect with disturbing accuracy. Especially since new cars already mostly have bluetooth anyway. Could probably be deployed as a firmware update to a not-insignificant portion of cars made after 2016 or so.
Skipping to self driving cars from that is kind of like saying "we built a house, therefore we can build a space elevator." Its a very different set of skills and scope.
Anecdotal, but I was recently at a summit where representatives from self driving car companies did a Q&A, and when they were asked about those sorts of infrastructure improvements they said that...
Anecdotal, but I was recently at a summit where representatives from self driving car companies did a Q&A, and when they were asked about those sorts of infrastructure improvements they said that they assumed that "smart city" infrastructure was 10+ years out, if it was ever going to be built.
I don’t know what those speeds are and how they compare to regular highway speeds, but I’d just like to slightly push back against it. I’ve been driving for a bit under two decades, and I’ve never...
One other I'd like to see NHTSA is to have mandatory governors that put a max speed
I don’t know what those speeds are and how they compare to regular highway speeds, but I’d just like to slightly push back against it.
I’ve been driving for a bit under two decades, and I’ve never got a speeding ticket. I very early on decided I never want to speed just for personal convenience, so to the best of my knowledge, I’ve only ever driven considerably above the speed limit in two or three occasions. All of them were for the safety of myself and other road users, and total combined duration speeding is probably less than 60 seconds (probably half that). They’ve all been on freeways or rural highways, nowhere near pedestrians or buildings.
If my car physically did not let me speed in those circumstances, I likely would have got into an accident (at least one head-on collision at highway speeds) and may not have been here alive today to write this comment.
Having said that: I’m all for a speed limiter that allows for brief bursts of high speed, and only kicks in for example if the vehicle has built up an average-above-speed-limit that shows disregard for speed limits across a few minutes, rather than just a rare speed-to-avoid-accidents.
Comment box Scope: comment response, information/history, personal reaction Tone: neutral, a little grim Opinion: yes Sarcasm/humor: none After a particularly horrific collision caused by people...
Comment box
Scope: comment response, information/history, personal reaction
Tone: neutral, a little grim
Opinion: yes
Sarcasm/humor: none
After a particularly horrific collision caused by people driving too fast, the National Traffic Safety Board (NTSB) recommended that 17 automobile manufacturers install "passive" speed governing instruments on their vehicles, and that the National Highway Traffic Safety Administration (NHTSA) mandate this in some way. These don't actively stop cars from going fast, but at least clearly alert the driver that they're speeding. They also recommended that the Insurance Institute for Highway Safety (IIHS) study the matter.
The NHTSA is certainly thinking about it. They will probably issue some sort of nothing-regulation that requires cars to beep slightly louder if you're speeding, but that could be followed by more stringent legislation.
Informally, people recognize the utility of two distinct kinds of speed governors:
Soft limit governors: making it difficult or annoying to exceed the posted speed limit, at least for brief periods, but not impossible
Hard limit governors: all the features of a soft governor, but also making it impossible to exceed the speed limit by a large margin for any amount of time, or by a small margin for a long period of time
These aren't formal definitions, but the way most people think about driving is "it's okay to speed a little bit sometimes," so our conversations tend to be distilled into the "sometimes" rules and the "all the time" rules. For example, most people would agree that it's okay to go 5mph, 10mph, or maybe even 20mph over the speed limit sometimes. But basically never 30mph, 40mph, 50mph over.
So-called "intelligent speed assistance" (ISA) is the technical way these limits are implemented. The technology isn't super well-developed. But the EU already has this sort of legislation on the books. There are two kinds of ISA tech, broadly:
Passive ISA: any technology that can identify when the vehicle is going too fast and alert the driver
Active ISA: any technology that is capable of slowing a vehicle in any way, like mechanically requiring more force to press the pedal down to go fast, or electronically preventing the motor from going too fast
These mostly map onto the soft/hard limit distinction that we would think about colloquially. The "passive" kinds of governors that the NTSB has recommended in the US are the "soft limit" kind. They don't really do anything, but they might tell you something. The NTSB hasn't recommended active ISA to enforce "hard limits."
It's possible to develop active ISA systems that the driver can override. These systems are still "active" and they still have limits. But there are mechanisms for the driver to get past them. You can implement this in any number of ways. One way might be to make it very physically challenging to push the pedal far enough to exceed the speed limit (i.e. dynamically change the pressure required to go further), but you can go fast if you seriously floor it. Another way might be for the active ISA not to even trigger until a certain amount of time has passed, say 15 straight seconds of speeding. Another way might be for active ISA to trigger immediately, but the car has some sort of switch or voice command saying "let me go fast for a minute" that lets you speed briefly. The front-end can kind of be whatever regulators/manufacturers want.
FMCSA has said its new proposal will require motor carriers operating trucks equipped with an electronic engine control unit (ECU) capable of governing the truck’s maximum speed to limit the truck to a speed as determined by the rulemaking and maintain that ECU setting for the service life of the truck.
The industry opposes most regulation. I sort of doubt any hard limits will go through, but it's possible. More likely, regulatory agencies will implement increasingly restrictive soft limits on any reasonable speeds, and hard limits only on extremely high speeds (like 100mph+), which industry doesn't reach anyway and wouldn't really care about.
I find it a little strange how fussed people get about this on the basis of privacy, although I find it especially strange how opposed commercial actors are to speed limits in general. Well, they have a profit motivation to go faster than is safe, because some actuary has calculated that their insurance will pay for all the damages for the pedestrians they kill, so maybe it's not strange. But it is definitely messed up.
I'm reminded of the quote from Benjamin Franklin: "Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety." And I'm also reminded that the context in which it was said has the exact opposite meaning:
The “essential liberty” to which Franklin referred was not what we would think of today as civil liberties but, rather, the right of self-governance of a legislature in the interests of collective security.
Replace the invading French with the SUVs of the 21st century, the wealth-focused Penn family with the powerful auto companies and lobbyists, the vetoing governor with automobile industry-funded politicians of today, and the colonial legislature with pro-traffic safety politicians, advocates, and constituents. Ostensibly weird mapping, but I actually think it holds in its original sense. I feel like pedestrians (100% of the population) collectively have more of a right to live than a smaller subset of drivers (<100% of the population) have a right to drive fast, in the same way that the legislature (which nominally represents the population) has more of a right to protect itself than a single powerful family has the right to not be taxed.
I'd rather have a solution that isn't pointlessly invasive. Personal privacy is a real issue. But I think being alive has to be the main goal here, not the lesser case of being able to occasionally use my phone while operating heavy machinery. And as far as commercial actors go, I feel like they should have less discretion than everyday people. It's an industry.
If my car physically did not let me speed in those circumstances, I likely would have got into an accident (at least one head-on collision at highway speeds) and may not have been here alive today to write this comment.
I'm glad you're still with us!
Statistically, there are few situations where going faster reduces the chance of fatality in a collision. The faster you go, the more momentum the vehicle has, and so the more force is being exerted on the vehicle in a potential collision. In general, permitting very high speeds to make an edge case safer results in more collisions across the population, who are continuing to drive at higher speeds in situations where it definitely does not make them safer. Even if speeding lets you escape a collision in a rapid-fire emergency situation, it increases the chance that the driver loses control of the vehicle (I struggle to think of any exceptions).
I tend to prefer infrastructure solutions to this problem because it's a lot harder for a bollard to malfunction than a piece of software. You can't hack a jersey barrier. It's trivial to design crash-resident infrastructure, but it's pretty hard to do this active stuff. But I think it's still worthwhile in some sense.
I completely agree with this. All the situations I was in, it was an unusual compounding of factors, and thinking back on it, I think having three of these incidents in less than 20 years of...
Statistically, there are few situations where going faster reduces the chance of fatality in a collision. The faster you go, the more momentum the vehicle has, and so the more force is being exerted on the vehicle in a potential collision.
I completely agree with this. All the situations I was in, it was an unusual compounding of factors, and thinking back on it, I think having three of these incidents in less than 20 years of driving places me well above the statistical average. I spoke to my dad about them and he reckons he’s been in three or four incidents like this, but across his nearly triple the number of years driving than I have.
It’s also worth noting that I think in all these cases, having access to higher acceleration could have just as easily pulled me to safety without going above the speed limit, but I’ve only ever driven “sensible”/budget cars that are fuel efficient but not very powerful. So a manoeuvre that took me 5-10 seconds to complete might have taken a sports car 2-3 seconds, if it was powerful enough.
haha this policy is so insane. I'm struggling to understand why they even bothered to announce it. "We have introduced a new standard! By our own analysis it will have no effect"
haha this policy is so insane. I'm struggling to understand why they even bothered to announce it.
"We have introduced a new standard! By our own analysis it will have no effect"
Comment box Scope: comment response, personal reaction Tone: gently corrective, but in agreement Opinion: yes Sarcasm/humor: none 67 annual deaths is not "no effect." That's 67 families every year...
Comment box
Scope: comment response, personal reaction
Tone: gently corrective, but in agreement
Opinion: yes
Sarcasm/humor: none
67 annual deaths is not "no effect." That's 67 families every year who don't have to grieve at the completely unnecessary death of a loved one. We get so caught up in numbers that we forget the realness of any change.
This regulation better integrates pedestrians into their testing procedures:
New Federal Motor Vehicle Safety Standard No. 228, Pedestrian head protection, would apply to passenger cars, light trucks (including pickups), multipurpose passenger vehicles (MPVs) (MPVs include sport utility vehicles (SUVs), crossover vehicles and vans) and buses with a GVWR of 4,536 kg (10,000 lb) or less. The standard would require vehicles to meet a head injury criterion (HIC) when subjected to testing simulating a head-to-hood impact. The vehicles would have to reduce the risk of serious to fatal head injury to child and adult pedestrians in impacts at vehicle speeds up to 40 km/h (25 mph), which encompass about 70 percent of pedestrian injuries from vehicle impacts. Moreover, it is expected the standard would be beneficial even at higher speeds.
These procedures are extraordinarily laborious to develop, which is why it takes them absolutely forever to issue rules. They have to develop really comprehensive criteria to evaluate safety, and it's actually not that straightforward to measure quantitatively. While they can use some technology and methodologies from other countries with better regulations (and this one does involve international consultation), the vehicle profile in the US is different, so it doesn't map 1:1.
The NHTSA can and should do better, but it isn't effective for them to solve all vehicle-related problems in one magical regulation. The way these rules are implemented means gigantic changes will attract enormous negative attention from auto lobbyists, will necessarily have complex, time-consuming public comment periods, and are likely to be pared down. It's usually more efficient to have a lot of smallish rule proposals in parallel, which is what they do. This is one such small proposal.
But the pace of the regulations is absurdly slow. (They don't always make press releases about new rules, so you have to go to the Federal Register to find some of them, like this one about automatic braking, but it's still slow.) They should definitely be taking more aggressive steps to reduce the danger vehicles pose to pedestrians. I think that if this rule passes, it will be followed by more like it... but it will take a long time. As advocates, we should be regularly calling on our elected officials to demand the NHTSA do more. Bureaucrats do have an incentive to listen to policy-makers, because the latter sets their budgets. The NHTSA could write tougher regulations if that was a specific goal of the current administration.
They also mention that the mandatory safety standard (FMVSS) will inform and coincide with an updated customer-facing safety ratings (NCAP):
Second, the standard would provide a regulatory counterpart to NHTSA’s planned crashworthiness pedestrian protection testing program in the New Car Assessment Program (NCAP) in the near term. On May 26, 2023, NHTSA published an NCAP Request for Comment (NCAP RFC) proposing to adopt a crashworthiness pedestrian protection program into NHTSA’s NCAP. NCAP would build on proposed FMVSS No. 228 and incorporate enhanced crashworthiness tests into NCAP that go beyond the specifications of proposed FMVSS No. 228.
Technically a safety rating doesn't reduce deaths (it's done after vehicle design and manufacturing), but consumers don't like buying cars with low safety ratings, so it provides a small financial incentive for automakers to prioritize pedestrian safety.
Congress could also easily write legislation protecting pedestrians, but they choose not to. The NHTSA and other organizations would still have to write the low-level rules, which could take years, but Congress could certainly expedite it. Currently, traffic safety is not an issue that most Americans understand or care about, but that can change through better activist messaging. People have become a lot more aware of this problem in the last 10 years, and especially the last 5 years, through urbanism's increasing digital presence.
The National Highway Traffic Safety Administration is advancing pedestrian safety by proposing a new rule to reduce fatalities and serious injuries among pedestrians struck by vehicles. The proposed rule would establish a new Federal Motor Vehicle Safety Standard requiring new passenger vehicles be designed to reduce the risk of serious-to-fatal injuries in child and adult pedestrian crashes.
The more crap they add to new vehicles the more money I'm willing to invest in keeping my three 15 year old beaters running for as long as possible. Seems like another regulation to increase cost,...
The more crap they add to new vehicles the more money I'm willing to invest in keeping my three 15 year old beaters running for as long as possible. Seems like another regulation to increase cost, homogeneity in design, and (should) only impact people who regularly drive in populated areas with frequent pedestrian interactions.
While I'm all for reducing pedestrian deaths, this seems underwhelming.
Based on their own numbers, this could save 67 lives a year. In 2022, there were 7522 deaths, up from a low in 2009 of 4109 deaths (source). Is there no other change that they could propose that would reduce fatalities by more than 1%? Given the ever increasing number of pedestrian deaths each year, will a 1% change even show up in the data?
Pedestrian deaths are largely due to a combination of infrastructure, distracted driving, and increased vehicle sizes. The NHTSA has the power to force the industry to change significantly. Make vehicles smaller. Make vehicles spy on us to make sure we're not looking at your phone. Do something.
If this were anything other than the auto industry, people would be angry about it. 7522 deaths per year is a lot of bodies. And that pile is only getting larger.
Absolutely against this wholesale. We have enough of a surveillance state already, and this would be abused immediately.
There's no reason this can't be done fully locally in the car.
Driving a multi-ton machine that can easily kill others means that you have restrictions placed upon you. This has always been part of the contract you make with the state to be allowed to drive a car. For example, you can be required to have blood drawn for testing if you are suspected of drunk driving, something that is infinitely more invasive.
But infinitely more infrequent and requiring probable cause, and only rarely done without consent.
I'd not trust the car to be able to actually see where I'm looking, in sunglasses, in my regular glasses, or for folks with different skin tones or eye shapes, historically we're not great at this.
If we go this route, out a breathalyzer lock on every car too. No driving if you have any alcohol in your system.
I'm not arguing for distracted driving or drunk driving but I do think this is way less reasonable than demanding safer cars.
In theory, such a system could just check to see if you're holding a phone (or anything) in your hand, whether or not you're looking at it. That would cover about 95% of distracted drivers today and it's simpler and less error-prone than eye-tracking.
That's actually already in the works - mandatory in every new car starting in 2026.
Holding something in my hand is gonna have a lot of false positives too.
And, much as I'm not sold on breathalyzer vehicles, they don't seem like they'd actually lock you out, it's just a check before you drive for your information. Hate it from a "clearly it'll be subpoenaed" standpoint. Maybe useful from a "I didn't think I was buzzed" standpoint. Not that useful from an "intentionally drunk driving" standpoint.
So what's the car gonna do, beep at me when I grab a sip of my drink, or put my hand on my thigh, or wave my arm around singing on a road trip? It'll get turned off so fast by anyone who can read a manual.
Once again, I almost never drink and am wary about driving after half a cider at a restaurant. I do work very hard not to get distracted by my phone, using audiobooks/podcasts and navigation on every trip to keep my brain busy and able to focus. I also haven't held my phone in my hand while driving for probably a decade now.
Anyway I'm not actually that pressed, because I don't think the tech is manageable, but I think it's the wrong focus for pedestrian safety contrasted with speed and vehicle size.
This will not happen.
This is
Bad driving, including driving under the influence, can be easily prevented without having to resort to a surveillance state. Advocating for one to be implemented because you're afraid of other people's driving is insane.
There’s no reason an Surveillance state needs to be horrifying. It’s still not even close to worth the risk.
Edit:
Now that I have some more time, hell there's LOTS of reasons that doing it locally in the car is a fucking problem.
What exactly does the system do on a positive identification of a phone use in the car? Pull the vehicle over? Stop it? Auto send a rate increase to your insurance?
How in the HELL do you control for false positives? The image detection software i've seen is impressive. It's not, "risk my life or my future" impressive.
How many people are you willing to punish? Shall we include all distracted driving? Single mother who took her eyes off the road to give her kid a bottle? Parent rushing to their child while on the phone as they're having some sort of major incident?
How much money are you willing to force everyone to spend? Cars are more and more becoming computers on wheels, but this sort of real time detection system, fully local, is going to cost a boatload more than any current system.
How much are you willing to punish drivers who break or mess with the system? You can drive with all sorts of damage to your vehicle, arguably legally depending on what kind. Am I unable to drive my car in an emergency because my detection system is damaged or broken? Oops a lens cracked or got covered so now i'm unable to drive? Or do you allow it and now you've just got a bunch of false positives again or people who already can barely afford to hang on getting punished even worse?
There's all sorts of grey area's in this that you're just asking to fight with. These idea's are always half baked and dangerous before you even get to the "and the state will 100% abuse this" scenarios.
This is already solved by existing partial self-driving technologies: nudge the wheel a bit to test the driver's attention, and if they don't respond in time, slow the vehicle down and pull over. As a bonus, this would save the lives of drivers who are having a medical emergency as well.
Test the driver's attention subtly.
I think I've addressed this above but I want to note that driving is the only place where this cavalier attitude to safety exists. I may just have been lucky, but I've never been in any job site or warehouse where people operating heavy machinery gave it anything but their full attention. But cars are styled like toys, not heavy machinery, and driver licensing in this country is a complete joke, so people don't take them seriously like they should.
What you're describing here is known as a "normalization of deviance". This is a classic human fallacy. Our attitudes to driving would not pass muster anywhere that safety is taken seriously.
Even if it didn't fit inside existing systems, I'd expect the overall cost of this kind of system to be < $50.
We already have this mostly figured out with our emissions systems. Something like 2-5% insist on bypassing regulations, but having the vast majority of people leaving things alone is a huge improvement.
This is before you get into cases where the car refuses to drive and someone dies because of it.
That's a reasonable position, but there's a good chance that in the near future our cars are going to spy on us anyway. Might as well put that to good use.
No. That's how these things always happen. Every single advancement of mass surveillance and erosion of rights is excused by some appeal to how something bad can be lessened. And the plan is always this nebulous.
It needs to be a criminal act to record people without reasonable consent, with a similar test to the GDPR: you can't make acceptance of the invasion of privacy a requirement to use a product that doesn't reasonably need that to function. And that goes doubly for government.
If you want to reduce pedestrian fatalities, one obvious thing has changed between 2013 and the present, in the period where pedestrian fatalities rose almost 60%: the average car will strike you in the upper chest and drive over you instead of hitting your knees and causing you to fall onto the hood. That's the thing that needs to be corrected.
I would much rather fight that trend than just lay down and take it.
If you want something concrete, in California Senate Bill 961 (To require passive speeding notifications in cars) is currently on the Governor's desk. If you're a resident, calling and asking him to sign would be very helpful. If it does pass, because California is such a significant market it's likely to do a lot to shape the likelihood of nationwide adoption of similar measures in new cars.
I'm more and more in favor of automated ticketing on toll roads. I used to be very against it, feeling it too authoritarian, but I've come around because car ownership needs to be heavily regulated anyhow, because cars are large dangerous pieces of machinery.
For speeding on highways in particular, it's also much less likely to have context-dependent wiggle room, false positives, and unintended consequences than something like red-light cameras, simply because physics dictates the limits. You measure the time between point A and point B and calculate the fee based on how much over the speed limit you did on average. While this wouldn't eliminate speeders on the whole, it would certainly drive a large enough segment of drivers, professionals in particular, that the speeders would be forced to drive the speed limit courtesy of congestion.
Automatic speed cameras are another thing happening in the Bay Area, though it's not mostly on highways. They're part of a package of improvements to help reduce speeding around pedestrians, as most of the vehicle-related injuries and deaths aren't from highway traffic, but in places where people trying to travel in various ways intersect. Cars are at their safest when everyone is moving in parallel lines with minimal variation in capabilities.
Oh the highway thing is beyond just reducing deaths, but also emissions. It tends to bleed in whenever I think about these things...especially since the necessary infrastructure is already completely in place on many highways.
Additionally, while it may be feasible in the Bay area, I can think of immense difficulty elsewhere having automated speeding tickets provide any tangible benefit without being overbearing to the point that people pass laws to ban them. You'd have to have checkpoints almost every intersection in a grid to be effective and that borders closer to 'tracking your every move from start to finish' than 'enforcing a baseline rule on major arteries.' If anything, having automated speeding in some places and not others off-highway could have adverse effects on passenger deaths as speeders try to figure out routes to avoid automatic ticketing. Why take the 45 mph road with automated speeding detection when you can blow through a 25mph residential area at 60 mph and be less likely to get caught?
However, I do think that killing speeding on highways will also reduce deaths off-highways, as conditioning people to obey the speed limit on highways will carry over to not-speeding everywhere.
While I have not dug in, I'd say that there are almost certainly few side benefits that will come along for the ride. Cars will almost certainly start trending downward in size....reducing the weight of a vehicle will be far easier to come into compliance than most other methods. This has added benefits of reducing wear on roads and reducing non-pedestrian deaths as well. One other I'd like to see NHTSA is to have mandatory governors that put a max speed of 70mph on car-exclusive roads like highways, and a max speed of 45mph on any mixed-use road. Or perhaps a mandatory-stop function when crosswalks are active. I very much like that one....all crosswalks emit a powerful signal when active, and if the car detects that signal at a ~10ft range engages emergency braking. There would be a lot of fender benders to start, but driver behavior would adjust pretty rapidly with that one.
While this isn't directly addressable by NHTSA, states have the most power to force drivers to chose smaller vehicles. One of the best rules we could get to trend vehicle size downward rapidly is increasing licensing difficulty depending on fully-laden GVW. Split off consumer vehicles into more classes, with more stringent road testing to be permitted to drive larger vehicles. This would allow light-vehicle licenses to have longer expiration duration and 0-cost, while heavy-vehicle licenses need more regular road tests. Having this clear separation could also allow increasing fines for moving violations based on the highest-weight vehicle you are licensed to drive. Having a speeding ticket (or run stoplight) be $20 if you're only authorized up to 1t vehicle, but $5000 for a 3t vehicle would force demand for smaller vehicles quite quickly.
If actual infrastructure is being developed and installed to solve these problems, then they can just skip directly to self-driving cars. Better yet, use that money for improved public transit and intercity HSR.
I do like this suggestion. It's kind of insane that I took my driving test in a subcompact car and now I can legally drive a motorhome towing that same car, or drive a gigantic pickup truck with a 5th wheel hookup and tow a 30 foot RV trailer.
The thing is though, it would be somewhat trivial to design such a signaling mechanism that just wires in directly to the crosswalk light, possibly even in-line with the bulb itself. Doesn't really need to be more than a passive BLE beacon, which we have cellphones able to detect with disturbing accuracy. Especially since new cars already mostly have bluetooth anyway. Could probably be deployed as a firmware update to a not-insignificant portion of cars made after 2016 or so.
Skipping to self driving cars from that is kind of like saying "we built a house, therefore we can build a space elevator." Its a very different set of skills and scope.
Anecdotal, but I was recently at a summit where representatives from self driving car companies did a Q&A, and when they were asked about those sorts of infrastructure improvements they said that they assumed that "smart city" infrastructure was 10+ years out, if it was ever going to be built.
I don’t know what those speeds are and how they compare to regular highway speeds, but I’d just like to slightly push back against it.
I’ve been driving for a bit under two decades, and I’ve never got a speeding ticket. I very early on decided I never want to speed just for personal convenience, so to the best of my knowledge, I’ve only ever driven considerably above the speed limit in two or three occasions. All of them were for the safety of myself and other road users, and total combined duration speeding is probably less than 60 seconds (probably half that). They’ve all been on freeways or rural highways, nowhere near pedestrians or buildings.
If my car physically did not let me speed in those circumstances, I likely would have got into an accident (at least one head-on collision at highway speeds) and may not have been here alive today to write this comment.
Having said that: I’m all for a speed limiter that allows for brief bursts of high speed, and only kicks in for example if the vehicle has built up an average-above-speed-limit that shows disregard for speed limits across a few minutes, rather than just a rare speed-to-avoid-accidents.
Comment box
After a particularly horrific collision caused by people driving too fast, the National Traffic Safety Board (NTSB) recommended that 17 automobile manufacturers install "passive" speed governing instruments on their vehicles, and that the National Highway Traffic Safety Administration (NHTSA) mandate this in some way. These don't actively stop cars from going fast, but at least clearly alert the driver that they're speeding. They also recommended that the Insurance Institute for Highway Safety (IIHS) study the matter.
The NHTSA is certainly thinking about it. They will probably issue some sort of nothing-regulation that requires cars to beep slightly louder if you're speeding, but that could be followed by more stringent legislation.
Informally, people recognize the utility of two distinct kinds of speed governors:
These aren't formal definitions, but the way most people think about driving is "it's okay to speed a little bit sometimes," so our conversations tend to be distilled into the "sometimes" rules and the "all the time" rules. For example, most people would agree that it's okay to go 5mph, 10mph, or maybe even 20mph over the speed limit sometimes. But basically never 30mph, 40mph, 50mph over.
So-called "intelligent speed assistance" (ISA) is the technical way these limits are implemented. The technology isn't super well-developed. But the EU already has this sort of legislation on the books. There are two kinds of ISA tech, broadly:
These mostly map onto the soft/hard limit distinction that we would think about colloquially. The "passive" kinds of governors that the NTSB has recommended in the US are the "soft limit" kind. They don't really do anything, but they might tell you something. The NTSB hasn't recommended active ISA to enforce "hard limits."
It's possible to develop active ISA systems that the driver can override. These systems are still "active" and they still have limits. But there are mechanisms for the driver to get past them. You can implement this in any number of ways. One way might be to make it very physically challenging to push the pedal far enough to exceed the speed limit (i.e. dynamically change the pressure required to go further), but you can go fast if you seriously floor it. Another way might be for the active ISA not to even trigger until a certain amount of time has passed, say 15 straight seconds of speeding. Another way might be for active ISA to trigger immediately, but the car has some sort of switch or voice command saying "let me go fast for a minute" that lets you speed briefly. The front-end can kind of be whatever regulators/manufacturers want.
The Federal Motor Carrier Safety Administration (FMCSA) has had a plan to implement speed governors in commercial vehicles for a few years now. It will probably be proposed in 2025.
The industry opposes most regulation. I sort of doubt any hard limits will go through, but it's possible. More likely, regulatory agencies will implement increasingly restrictive soft limits on any reasonable speeds, and hard limits only on extremely high speeds (like 100mph+), which industry doesn't reach anyway and wouldn't really care about.
I find it a little strange how fussed people get about this on the basis of privacy, although I find it especially strange how opposed commercial actors are to speed limits in general. Well, they have a profit motivation to go faster than is safe, because some actuary has calculated that their insurance will pay for all the damages for the pedestrians they kill, so maybe it's not strange. But it is definitely messed up.
I'm reminded of the quote from Benjamin Franklin: "Those who would give up essential Liberty, to purchase a little temporary Safety, deserve neither Liberty nor Safety." And I'm also reminded that the context in which it was said has the exact opposite meaning:
Replace the invading French with the SUVs of the 21st century, the wealth-focused Penn family with the powerful auto companies and lobbyists, the vetoing governor with automobile industry-funded politicians of today, and the colonial legislature with pro-traffic safety politicians, advocates, and constituents. Ostensibly weird mapping, but I actually think it holds in its original sense. I feel like pedestrians (100% of the population) collectively have more of a right to live than a smaller subset of drivers (<100% of the population) have a right to drive fast, in the same way that the legislature (which nominally represents the population) has more of a right to protect itself than a single powerful family has the right to not be taxed.
I'd rather have a solution that isn't pointlessly invasive. Personal privacy is a real issue. But I think being alive has to be the main goal here, not the lesser case of being able to occasionally use my phone while operating heavy machinery. And as far as commercial actors go, I feel like they should have less discretion than everyday people. It's an industry.
I'm glad you're still with us!
Statistically, there are few situations where going faster reduces the chance of fatality in a collision. The faster you go, the more momentum the vehicle has, and so the more force is being exerted on the vehicle in a potential collision. In general, permitting very high speeds to make an edge case safer results in more collisions across the population, who are continuing to drive at higher speeds in situations where it definitely does not make them safer. Even if speeding lets you escape a collision in a rapid-fire emergency situation, it increases the chance that the driver loses control of the vehicle (I struggle to think of any exceptions).
I tend to prefer infrastructure solutions to this problem because it's a lot harder for a bollard to malfunction than a piece of software. You can't hack a jersey barrier. It's trivial to design crash-resident infrastructure, but it's pretty hard to do this active stuff. But I think it's still worthwhile in some sense.
Related: you can reduce speeding by 90% just by automatically enforcing speeding tickets though (with a camera, not a cop). The financial motivation is always compelling.
I completely agree with this. All the situations I was in, it was an unusual compounding of factors, and thinking back on it, I think having three of these incidents in less than 20 years of driving places me well above the statistical average. I spoke to my dad about them and he reckons he’s been in three or four incidents like this, but across his nearly triple the number of years driving than I have.
It’s also worth noting that I think in all these cases, having access to higher acceleration could have just as easily pulled me to safety without going above the speed limit, but I’ve only ever driven “sensible”/budget cars that are fuel efficient but not very powerful. So a manoeuvre that took me 5-10 seconds to complete might have taken a sports car 2-3 seconds, if it was powerful enough.
haha this policy is so insane. I'm struggling to understand why they even bothered to announce it.
"We have introduced a new standard! By our own analysis it will have no effect"
Comment box
67 annual deaths is not "no effect." That's 67 families every year who don't have to grieve at the completely unnecessary death of a loved one. We get so caught up in numbers that we forget the realness of any change.
This regulation better integrates pedestrians into their testing procedures:
These procedures are extraordinarily laborious to develop, which is why it takes them absolutely forever to issue rules. They have to develop really comprehensive criteria to evaluate safety, and it's actually not that straightforward to measure quantitatively. While they can use some technology and methodologies from other countries with better regulations (and this one does involve international consultation), the vehicle profile in the US is different, so it doesn't map 1:1.
The NHTSA can and should do better, but it isn't effective for them to solve all vehicle-related problems in one magical regulation. The way these rules are implemented means gigantic changes will attract enormous negative attention from auto lobbyists, will necessarily have complex, time-consuming public comment periods, and are likely to be pared down. It's usually more efficient to have a lot of smallish rule proposals in parallel, which is what they do. This is one such small proposal.
But the pace of the regulations is absurdly slow. (They don't always make press releases about new rules, so you have to go to the Federal Register to find some of them, like this one about automatic braking, but it's still slow.) They should definitely be taking more aggressive steps to reduce the danger vehicles pose to pedestrians. I think that if this rule passes, it will be followed by more like it... but it will take a long time. As advocates, we should be regularly calling on our elected officials to demand the NHTSA do more. Bureaucrats do have an incentive to listen to policy-makers, because the latter sets their budgets. The NHTSA could write tougher regulations if that was a specific goal of the current administration.
They also mention that the mandatory safety standard (FMVSS) will inform and coincide with an updated customer-facing safety ratings (NCAP):
Technically a safety rating doesn't reduce deaths (it's done after vehicle design and manufacturing), but consumers don't like buying cars with low safety ratings, so it provides a small financial incentive for automakers to prioritize pedestrian safety.
Congress could also easily write legislation protecting pedestrians, but they choose not to. The NHTSA and other organizations would still have to write the low-level rules, which could take years, but Congress could certainly expedite it. Currently, traffic safety is not an issue that most Americans understand or care about, but that can change through better activist messaging. People have become a lot more aware of this problem in the last 10 years, and especially the last 5 years, through urbanism's increasing digital presence.
The more crap they add to new vehicles the more money I'm willing to invest in keeping my three 15 year old beaters running for as long as possible. Seems like another regulation to increase cost, homogeneity in design, and (should) only impact people who regularly drive in populated areas with frequent pedestrian interactions.