Slashdot Mirror

Alcoholic crashes involve people falling asleep. It's weariness that is the problem (and, of course, for a small fraction of people, alcohol makes them weary).

Falling asleep (behind the wheel or otherwise) due to alcohol consumption only happens to someone who is really blasted. Your ability to drive is affected well before you drink that much. From this article:

Alcohol is a substance that reduces the function of the brain, impairing thinking, reasoning and muscle coordination. All these abilities are essential to operating a vehicle safely.

Although the driver definitely deserves criticism for her actions, this ruling is very convenient for Uber.

This means that they can relax significantly when it comes to the safety of the testing vehicles; if an accident happens, the courts says it's the driver's fault. Uber can sit back and blame the driver. Very handy.

The court should take into account the hazards posed to the public; if a company tests out new driving technology on public roads where the sole purpose is to disconnect the driver from the driving task, the hazards arising from the driver actually becoming disconnected should be taken into account. Especially, the automotive safety standard (ISO-26262) explicitly states that 'foreseeable misuse' MUST taken into account in the hazard analysis.

Is 'using the phone in the car while driving' a foreseeable misuse? You bet it is. We have all seen it (and done it). NHTSA has emphasized this in a recent report, saying that this is not only applicable for L3 and L4 AD systems, but even for L2 systems: https://www.nhtsa.gov/sites/nh...

So, if the hazard analysis identifies this risk, but the company neglects to handle it in such a way that someone gets killed, shouldn't it be held liable? I think so. If the company neglects to perform a proper hazard analysis (i.e. not using established, acknowledged standards and best practice) and therefore misses hazards and therefore have someone killed, shouldn't they be held liable? I think so.

Ergo; I think Uber needs to be held responsible for this.

Let's put it this way: the evidence for warming from CO2 is good. The sensitivity is nothing like the scenarios you suggest. The evidence for all the feedbacks needed to generate "runaway warming" is flimsy to non-existent.

Nobody claimed it was a runaway effect. The estimate is based on past, present and projected pollution. The report has 500 pages explaining this or you could have simply read the article.

There's been a 14% greening of the planet over the last 33 years as a result of the extra CO2 in the atmosphere (from satellite studies). This remarkable fact has been almost invisible to the mainstream media, NGOs and other activist scientists involved in perpetuating the paradigm.

I'm aware that the planet is getting greener as a result of CO2 and climate change. It's understood that flora will flourish in some regions while dying in others. The issue with climate change has always been about the death of fauna and the migration of arable regions of land. It's when the weather becomes erratic that crops are threatened but some flora will thrive. This doesn't mitigate the damage of climate change.

I would suggest you turn down your hysterics knob a few clicks.

You dismiss science as being hysterical instead of pointing to research that counters the evidence put forth? You response is as sound as that of an anti-vaxxer who is certain autism is a just a shot away.

Who cares about what a pathetic compulsive liar like Musk tweets?

Virtually all of his grandstanding tweet lies have been debunked.

Most recent debunked lie: "our cars have the highest NHTSA whatever".

https://www.nhtsa.gov/press-re...

https://www.nhtsa.gov/vehicle/... tells me that the model 3 isn't yet rated.

What's your fucking source?

You do realize that nearly all popular car models get a 5-star crash test rating? It's just that for some reason, a car getting a 5-star safety rating becomes a front-page news story only when it's a Tesla.

(This is not to denigrate Tesla's safety engineering. The lead safety engineer at BMW gave a guest lecture at my graduate structural engineering class. To our surprise, safety is dead last in the design process. First the artist designs the basic shape of the car. Then all the engineers design the mechanical and electrical components to fit within the artist-defined shape. Finally the safety engineer is given a budget of (say) 25 kg of steel, and told to add reinforcement to make the car pass the government and insurance institute safety tests. So all it would take to design a safer car is to move the safety engineer earlier in the design process, which I believe Tesla has done.)

... the level of THC in the blood and the degree of impairment do not appear to be closely related.

You may be interested in this NHSTA study, which found drugs had NOTHING to do with accidents.

The premise is simply false. Tesla does not have a high rate of recalls. Go here. Punch in "2017 Tesla". Check out the recalls and investigations stats. Now punch in "2017 Mercedes". Note the difference.

Anybody who read the NHTSA report should clearly understand that the Autopilot safety data comparison was not done to demonstrate the safety of Autopilot, but rather to decide if there was indication that AP caused an increase. Also, 2/3 of the cars in the study didn't have any pre-AP data at all. It was entirely useless for the purpose of making any kind of safety claim. The NHTSA should not have had to clarify, but too many idiots made stupid claims based on that information. The media in general can be really stupid with statistics.

One thing though,

I thought those number were about Tesla autopilot on *almost ideal condition* VS people on *all condition* no?

https://static.nhtsa.gov/odi/i...

You keep trying to side track the conversation.

Says he of the ever moving goalposts. Waaay back in this conversation, you said

So this is the version where they got it not to kill people? Funny, I would have put that feature in first.

You didn't say "better than humans" or any sort of comparison to humans. You effectively said, "It has to be perfect."

As soon as I pointed out how perfect is a ridiculous standard, considering humans aren't perfect, you started shifting those goalposts. Next it was, "why couldn't the system see something people can see?". What you've done is pick one situation where the system didn't work as intended, you ignored hundreds of the exact same situation, in the exact same part of the exact same road where it did, so I guess it makes sense for this to be an issue for you. Rational people don't do that. Only people with a serious agenda cherry pick like that and ignore all the evidence to the contrary.

In your last two sentences, you've finally gotten to a reasonable argument to have:

Human driving is very safe if you consider the 3.2 *trillion* miles driving in the US every year. Much safer than Autopilot would be if it were to drive everywhere and in all conditions like a human.

Now we're comparing rates of accidents between humans as a whole and autopilot as a whole. Bravo, this is actually meaningful, and what I was saying all along, albeit indirectly.

So lets have that discussion. There isn't a ton of research on this, but there is a little. AUTOMATED VEHICLE CRASH RATE COMPARISON USING NATURALISTIC DATA is from 2016 and finds that self driving cars are a lot safer. How Safe Are Self-Driving Cars? Waymo Proves They’re Pretty Darn Safeis from 2017 and summaries WayMo data, which you can argue is likely a bit biased towards Google. But even in the worst case scenario, they find self-driving cars an order of magnitude safer than human drivers.

In the NHSTA's report on the Tesla crash where it hit the semi-truck side-on and took off the top of the car, they interestingly don't find fault with Tesla. Just that they needed to be more specific on the limitations of autopilot. And buried in that report is a graph that shows auto-steer dropping the accident rate (airbag deployment) per million miles from 1.3 to 0.8 when it's enabled. I'm finding it hard to find non-fatality rates for the whole US, but in the couple of states I looked at, the accident rate seems to be between 1 and 3 for most types of roads.

If you've got more comprehensive and current research, I'd be very happy to see it. Unfortunately there doesn't seem to be a ton, but what's out there seems to strongly suggest that the current self-driving tech is at least as safe as the average driver, if not more safe, in most circumstances.

Here is the full rule (1500 pages) for 2012-2016 if you'd like to read it, but I'll summarize a bit for you.

https://www.nhtsa.gov/staticfi...

> fleet-wide averages, without as much exception for "trucks."

There are two (or more) completely separate fleets. Cars, light trucks, medium trucks (and busses), heavy trucks, motorcycles. There is no "exception", the two groups are computed entirely separately, based on entirely different MPG standards and different average lifetime miles.

For CAFE purposes, each company is essentially split into two companies - a truck company and a car company. (Also motorcycles and large trucks are computed entirely separate, as if they were different companies). You can read the full details in the EPA rules above.

So first the company does its cars. The first step on calculating the car standard is to find the average size (footprint) of the company's cars. I'll directly quote the EPA rule on this rather than trying to explain it in my own words:
--
EPAâ(TM)s final standards, like the standards NHTSA
promulgated in March 2009 for MY 2011, are expressed as mathematical functions depending on vehicle footprint. Footprint is one measure of vehicle size, and is
determined by multiplying the vehicleâ(TM)s wheelbase by the vehicleâ(TM)s average track width.
--

After finding the footprint, you look at the table (section 3, I think) that gives the formula for your range. Inputting the average footprint, the formula tells what the average fuel economy needs to be, in GALLONS PER MILE.

It's gallons per mile because a vehicle that gets 1MPG burns twice as much gas as one that gets 2MPG, but a vehicle that gets 99MPG is almost the same as one that gets 100MPG.

Subtract your company's ACTUAL average GPM for cars from the standard to get the amount of credit or debit. If the company is more efficient than required, it can either save those credits for next year, or sell the credits to another car company. Similarly, if this year's sales aren't efficient enough, the company can either use credits it earned in an earlier year, or buy credits from a more efficient company. (Credit brokers are allowed, but cannot actually own the credits, only bank them).

Once your done with the cars, you go through the same procedure, separately, for your motorcycles, then again completely separately for light trucks, etc.

I mentioned that a company that doesn't meet its target can buy credits from a company that the target. What Mack beats their heavy truck target, while BMW needs to buy credits for their cars? Mack has truck credits to sell, BMW wants to buy car credits. The public doesn't care whether a gallon of gas is burned in a motorcycle or a bus, they only care how much as is burned, so before trading companies can apply a formula to convert light truck credits to car credits, or car credits to medium truck credits or whatever. (It's not one-for-one, different kinds of credits are "worth" different amounts). Note that it may be Volvo's truck credits offsetting Ferrari's car debit. The Corporate in CAFE doesn't matter once you start trading different kinds of credits.

Just as GMC can convert truck credits to (fewer) car credits in order to sell them to Ferrari, GMC can also convert truck credits to car credits for Buick. GMC and Buick happen to be the same company, but GMC could just as easily trade those credits to a different company, maybe Ford or Volkswagen.

Again, the full details are in the actual rule linked above, but the summary is that car, light truck, medium truck, and heavy truck are computed completely separate, like separate companies. There is no averaging between cars and trucks.

> Most cars won't have self-driving capability or car to car communication.

Doesn't take most, just takes enough. Vehicle to Vehicle (V2V) was introduced to new cars 2 years ago, and will likely be in all news cars in 2 years. Driver assist (lane assist, emergency braking...) with NHTSA is already recommending CIB, DBS and PAEB braking, and was on a path to make them mandatory, but the current anti regulation path in the US has likely put a delay to that.

With these style of cars, we are a few system updates away from converting them to be part of a smart grid. So 15 years may be optimistic with a deregulation movement likely to slow things for the next 3-7 years. But if we had a hard regulatory push away from petrol cars in 10 years, we could be 90% smart cars in 15-20 years.

Go here. Let's pick a relatively recent year, so it reflects relatively modern manufacturing, but not so modern that there won't be time for problems to come up. Say, 2015? So punch in, say, "2015 Tesla" in one year, then pick some other manufacture and do the same thing - I'll do "2015 BMW" or "2015 Mercedes". Let's see how many recalls come up for each of the models that come up.

Tesla: 1, 1, 1, 1, 1, 0, 0, 1
BMW: 2, 1, 1, 1, 1, 3, 3, 3, 3, 3, 1, 1, 3, 3, 3, 3, 3, 0, 0, 0, 3, 1, 0, 1, 2, 2, 2, 2, 2, 2, 2...
Mercedes: 7, 7, 7, 7, 7, 4, 1, 8, 8, 8, 8, 0, 0, 3, 3, 3, 3, 8, 8, 7, 7, 1, 1, 2, 2, 4, 4, 6, 6....

My god, look at that horrible Tesla build quality!

Furthermore, Tesla has - very unusually - never had a mandatory recall forced on them by the NHTSA, or one that started from a NHTSA investigation. Every single recall Tesla has ever had has come internally.

Now back in reality:
SpaceX reliability is right on average for the space industry even if you take into account early experimental failures. If you only count payloads lost it's better than average. They are beaten only by ULA, and only because of one single failure to deliver a payload.

Telsa's Autopilot according to the NHSTA drops the highway accident rate of these vehicles by 40% making an autopilot driven Tesla currently the safest way to move on the highway. https://static.nhtsa.gov/odi/i... (Figure 11)

Mind you if you did get into an accident you'll probably want to be in a Tesla given that most of the models are widely considered the safest cars on the road, and the Model S achieved a record high rating by the NHTSA and NCAP and the Model X was the only SUV to ever be awarded 5 stars by the NHTSA as well. https://www.nhtsa.gov/vehicle/...

As for the Boring company, no doubt its safety will be a story as boring as your lame post.

Now back in reality:
SpaceX reliability is right on average for the space industry even if you take into account early experimental failures. If you only count payloads lost it's better than average. They are beaten only by ULA, and only because of one single failure to deliver a payload.

Telsa's Autopilot according to the NHSTA drops the highway accident rate of these vehicles by 40% making an autopilot driven Tesla currently the safest way to move on the highway. https://static.nhtsa.gov/odi/i... (Figure 11)

Mind you if you did get into an accident you'll probably want to be in a Tesla given that most of the models are widely considered the safest cars on the road, and the Model S achieved a record high rating by the NHTSA and NCAP and the Model X was the only SUV to ever be awarded 5 stars by the NHTSA as well. https://www.nhtsa.gov/vehicle/...

As for the Boring company, no doubt its safety will be a story as boring as your lame post.

The United States already has a De-facto import duty against importing Chinese cars in that none so far are able to get https://www.nhtsa.gov/ approval for importing them into the United States.

You could save more lives by banning unhealthy junk food (Roughly as many people died of diabetes as traffic fatalities last year) and a whole load of other things.

False equivalence. Eating badly only kills you. Distracted driving risks the lives of other people. If you want to slowly kill yourself with a bad diet or drug habit go ahead. But when your actions start to threaten others then absolutely the state should intervene. That's the whole point of a government and to solve problems that we cannot solve ourselves.

I reject your premise that people's fears have any worth in determining rules.

False premise again. What fears? Distracted driving is proven to cause thousands of unnecessary and preventable deaths every single year. All because people aren't willing to delay gratification on their entertainment for a few minutes.

If people are afraid of terrorists attacking shopping malls, should we put TSA stations up at every entrance point?

Oh I get it. You want to get all hyperbolic with absurdities rather than actually point out a better solution. What is your solution to distracted driving? Or maybe you think it is fine. Is your argument is that your right to send/receive text messages while driving is more valuable to our society than the lives of several thousand people and injuries to tens of thousands?

NHTSA is counting

My state has been keeping distracted driving stats since before smartphones were commonly available.

And then there is the standard correlation does not mean causation.

I guess techie readers of slashdot are just as willing to fall into confirmation bias as the general public. The analysis in TFA is poor. But it supports what a lot of people here want to believe, so who cares?

Well, it's a lot different from listening to the radio, since music typically is just a background activity, and even if you are listening to news / talk, you don't have to respond. It's really a question of how much concentration is required to talk vs. send a text, and whether that concentration distracts you from your primary responsibility, which is to drive the car safely.

Uh, music is more than just a background activity to a lot of drivers. And a lot of people yell back at talk radio. Of course, you are too attentive to the road to notice what other drivers are doing, right?

Got a new car with a touch screen factory radio. It takes more attention to select music or a podcast through it than to do the same from my phone, but the radio is safer to use because, well, just because, right?

How about some numbers? In 2015, distracted driving was a factor in 10% of accidents and cell phone use was a factor in 14% of that 10%. But you wouldn't guess that based on most commenters here.

I thought those numbers sounded like bullshit, but apparently the info is from the NHTSA. (At least the 2500%).
Page 6: https://www.nhtsa.gov/staticfi...

Still, I think I'll wait a bit after that 6 pack of craft beer.

Your opinion does not agree with the conclusions of the HTSA report.
It looked at driver engegement and how it was affected by driver assisting features. Conclusion is that indeed some periode of inattentiveness exist but rarely bigger than 5 sec. So the 7 seconds in which the driver did not react to the truck crossing his path is very exceptional.
Secondly they looked at the amount of accidents and collisions of Tesla's before and after the Autopilot was introduced. They fell by 40 percent.

In my opinion a good attentive driver, even with automatic systems engaged, will still keep his attention where it belongs: on the road.

So this article links to a story here, and on the very same page, NHTSA links to the remote exploits story below. Someone there needs to connect the dots.

https://www.nhtsa.gov/press-re...

Motor Vehicles Increasingly Vulnerable to Remote Exploits
https://www.ic3.gov/media/2016...

I've no idea what those 15 points are...

Here you are 15 points check. I googled it for you. ;)

* Drunk driving crashes continue to represent roughly one-third of fatalities, resulting in 9,967 deaths in 2014.
* Distracted driving accounted for 10 percent of all crash fatalities, killing 3,179 people in 2014.
* Drowsy driving accounted for 2.6 percent of all crash fatalities; at least 846 people died in these crashes in 2014.

Autopilot killed 1 person; ever. And it did so on a location where it was not intended to be used. The safety record doesn't mean nothing, but there is still room for improvement. The difference is now the improvement can be shared to all cars via an update. In a standard fatal accident, the improvements are much harder to act upon. Reference: http://www.nhtsa.gov/About+NHT...

"Part of the reason why cars are safer today than 40 years ago is because the CDC did studies examining car safety and provided the results and recommendations to Congress."

That's a red herring. The CDC shouldn't be involved there either, they should be focused on preventing the zombie apocalypse. Vehicle safety should fall under the NHTSA. And workplace safety should fall under OSHA. But, for political reasons, the CDC duplicates the work of both.

This isn't something I can imagine doing anywhere near as badly in a manual. You panic, you stomp brake and clutch. Miss the brake and go for the accelerator, and you rev like crap but don't accelerate. You miss the clutch, you stall it. Seems like quite a challenge to miss the clutch and hit the foot rest, whilst simultaneously missing the brake and hitting the accelerator.

http://www.nhtsa.gov/staticfil...

Researchers reviewed each crash narrative to determine whether the crash actually resulted
from a pedal application error. Of the 2,930 crashes, 2,411 were caused by a driver applying the
accelerator when he or she intended to apply the brake. Fifty-eight were the result of the driver’s
foot slipping from the brake and pressing the accelerator, 47 were the result of the driver pressing
the wrong pedal in a vehicle with manual transmission (either clutch or accelerator rather than the
brake, or the brake rather than the clutch). Reviewers determined the remaining 414 crashes not to
11 be the resultt of a pedal misapplication; these 519 incidents were therefore excluded from the present
analyses.

Do you quiz car dealerships on where and under what conditions the raw materials of their vehicles were mined?

You may want to inform yourself of the American Automobile Labeling Act.

'Apply pressure then let off...'. Hmm, seems like a heard that somewhere else with a different name. Now what was it? Oh, yeah, PUMPING THE BRAKES.

Pumping the brakes has never meant 'stomp on then completely release the brakes'. It means to modulate the pressure so as to prevent lockup.

Here is what the NHTSA says ABS is:

What ABS does is similar to a person pumping the brakes. It automatically changes the pressure in your car's brake lines to maintain maximum brake performance just short of locking up the wheels. ABS does this very rapidly with electronics.

Emphasis mine.

I'm telling you what the traffic safety experts recommend after years of study.

The 85th percentile rule is based on a study done in 1964 by David Solomon, back when car crashes were much less livable and people had drive more carefully. Have any studies been done recently to demonstrate that the 85th percentile rule is still pertinent today?

I'd like to look at the data. Could you provide a link to this information?

Go to http://www.nhtsa.gov/FARS and run 1000 or so search queries on every topic you want.

Instead of telling me to google it, could you point to a specific study?

And all the farmers driving their tractors on the road. And bicyclists. And the Amish.

All of those are banned from the interstate. They are limited to slower and smaller roads, ones appropriate to that type of vehicle.

Do you think the fastest 15% of the people on the road (the speeders) should determine the speed limit? That sounds like an extremely bad idea to me.

I'm telling you what the traffic safety experts recommend after years of study. That you object to reality will not change it.

Common sense is rarely right when applied to a subject you don't have actual knowledge of. That you think it's a bad idea is proof that you are both ignorant of the topic, and arrogant about your ignorance at the same time. A lethal combination.

I'd like to look at the data. Could you provide a link to this information?

You obviously have no idea what you are talking about. Where to start? Go to http://www.nhtsa.gov/FARS and run 1000 or so search queries on every topic you want. Most of the "studies" are re-statements or analysis of FARS data, as it's one of the most complete databases of its kind in the world (yes, much of the rest of the world builds local law based on US-only data, as it's the best source for the data). Then go to https://www.motorists.org/ and see what they have on how to set a proper speed limit. The NMA will have lots of cites, no need to repeat them all here for someone that's demanding "citation needed" as a dismissive, rather than an honest query. If it was and honest question, why did't you google https://www.google.com/search?... or https://www.google.com/search?... ?

The answer is, you don't actually want an answer, you just want to argue. When the two most obvious search strings I think of give first links to TXDoT and USDOT manuals recomending setting the limit at 85% for optimal safety, why would you question it? Where did you get your traffic engineering degree? TTI, as an engineering extension to Texas A&M is a good place to start.

There, cites, and hundreds of hours of work to educate yourself. Thousands of hours of work if your mind is as closed as it appears.

You fear complexity... but as I said, the numbers don't lie...

http://www.nhtsa.gov/About+NHT...

AEB technology is already showing benefits in the real world. Several studies, including a recent report from IIHS, show that AEB technology can reduce insurance injury claims by as much as 35 percent.

Could this technology hurt one person, somewhere, ever? Yes. But it will help far more than it hurts.

It is similar to airbags. Airbags have sometimes been a problem and actually hurt people by going off at the wrong time, but in total, they have saved FAR more lives than they have taken.

You fear the edge case while ignoring the majority of times when this would save someone from a crash. That is emotional reasoning, not sound logic based on facts.

NHTSA publishes a list of civil settlements here:
http://www.nhtsa.gov/Laws+&+Re...

Fiat Chrysler was recently fined for inadequate protections on Jeep gas tanks, but I did not see that on the page linked above - so the list isn't entirely current.

NHTSA may not be the fastest regulatory group out there, but they have shown a willingness to go after car companies that do not issue timely fixes for dangerous problems. Automotive software bugs will eventually kill people. Unfortunately, NHTSA probably won't care until then.

It's quite literally only a matter of time, eventually the 1987 Malibu will leave the roads.

So once one comes to accept that V2V will be implemented as a standard and one must also hope they don't fuck it up too bad, it will be capable of drastically improve vehicle safety.

When cars in the immediate vicinity can communicate their future intents, e.g.; I am turning left if 50 years, trailing vehicles can react accordingly.

Ultimately cars will automate, for safety it will be required in the same way that V2V is required. Non-unatomated cars will be detectable by the absence of V2V response and avoided. It's possible far enough into the future, they will be outlawed form public roads.

But cars will have HUDS that communicate what the vehicle knows to the person responsible for controlling the vehicle. Why the upgraded package, the cars will do so only moments before it reacts for you.

It could be argued that this HUD information may actually be required to help some people feel comfortable with the transition, to trust the vehicles.

While, others will gladly give over control to the car and pay for the extra features regardless of HUD.

What sane person who could afford it, wouldn't pay $5k, once, for a chauffeur? Assuming again, it's not fucked up by shitting engineering, that investment will last the length of ownership; the time savings will more than pay for the system.

http://www.statisticbrain.com/driving-citation-statistics/

The National Highway Traffic Safety Administration supports the 300K per officer number. Their site: http://www.nhtsa.gov/ is a standard impossible to use, ugly and cumbersome site, so digging for the data to back up their statistic will be a task for the reader. I can only imagine that they don't include all police officers in that statistic, and only include those that wrote at least 1 traffic ticket. After all Federal Officers would be considered sworn, but in few instances would have jurisdiction to right a ticket. Many State and Local Officers would have the jurisdiction, but job duties wouldn't put them in position to write traffic tickets.

No, those numbers have to be wrong. That would make only 20k officers in the US.
Just NYC has 40k officers. Granted some of them might not be traffic cops but the 300k
per officer is misleading at best but most likely flat out wrong.

Friction force is a function of the coefficient of friction times the downward force. The simple physics interpretation says that the stopping distance is independent of the "amount of rubber per pound":

http://en.wikipedia.org/wiki/B...

In practice, particularly with anti-lock brakes, the situation is a bit more complicated, but it shouldn't be as bad as double. I notice that you've chosen the "truck with hot brakes" versus the car, presumably with cool brakes, for your comparison. That's not only not a fair comparison, riding around in emergency braking situations with hot brakes all the time is the result of either poor maintenance or poor driving.

Note that US federal regulations require large trucks to stop in considerably less than twice the distance of passenger cars (or even motor bikes), and that tested trucks average quite a bit better than that maximum: http://www.nhtsa.gov/DOT/NHTSA....

So if a truck does take twice the distance of a car, it may literally be criminally poorly maintained.

What he said^^^^^^
Here is the relevant regulation quoted from the linked pdf.

"Current FMVSS No. 121 regulations allow longer stopping distances for pneumatically
braked heavy vehicles than for passenger cars or motorcycles. NHTSA believes that
improving the discrepancy in stopping distances is very important in reducing heavy
truck related fatalities in North America. Currently, pneumatically braked truck tractors
are required to stop from 96.6 kph (60 mph) in 108 m (355 ft.) at GVWR, whereas the
FMVSS No. 135 requirement for passenger cars is 65 m (216 ft.) Actual stopping
distances seen in the field vary significantly for both groups."

Also, another thing is that on wet pavement, truck and auto stopping distances are about the same.

so to suggest that the auto industry will follow some parallel of the PC industry is just silly.

Yep. Further: there are very few industries as overly-burdened with Federal requirements (see: http://www.nhtsa.gov/cars/rule... for just 1 example) as the auto industry is. The connection TFA makes between the two industries is tenuous at best. More accurately: it's non-existent.

Friction force is a function of the coefficient of friction times the downward force. The simple physics interpretation says that the stopping distance is independent of the "amount of rubber per pound":

http://en.wikipedia.org/wiki/B...

In practice, particularly with anti-lock brakes, the situation is a bit more complicated, but it shouldn't be as bad as double. I notice that you've chosen the "truck with hot brakes" versus the car, presumably with cool brakes, for your comparison. That's not only not a fair comparison, riding around in emergency braking situations with hot brakes all the time is the result of either poor maintenance or poor driving.

Note that US federal regulations require large trucks to stop in considerably less than twice the distance of passenger cars (or even motor bikes), and that tested trucks average quite a bit better than that maximum: http://www.nhtsa.gov/DOT/NHTSA....

So if a truck does take twice the distance of a car, it may literally be criminally poorly maintained.

They have a careers subpage. I would be willing to bet its got a form or two, and that's *very* personal info.

http://www.nhtsa.gov/Jobs has no login at all I could see. Most sites like that will deep-link to https://www.usajobs.gov/ which is secured-only. Seems to do pretty well at it today, but no reason to not turn on SSL for the sites with no personal information.

There's almost no reason to not default to HTTP for everything. There's no reason to not just encrypt it, even for static pages.

There is absolutely no reason to use HTTP for anything. Encrypting the connection costs very little, prevents you from having stupid mistakes by not encrypting things that need to be, and provides enhanced privacy to things you may not realize that person is sensitive on. There's no reason NOT to make HTTPS everywhere.

Yup, that's what I said. The only reason not to is if you have a very popular web site with only static content. SSL on that will drain resources for minimal gain.

http://www.nhtsa.gov/

I couldn't find a login there, though it doesn't have any personal information or data on it. Just statistics and articles. But putting in your email to subscribe is in plaintext.

There's almost no reason to not default to HTTP for everything. There's no reason to not just encrypt it, even for static pages. No real benefit, but no real loss either.

In order to translate "lives lost" into money, you need to start by assigning a value to a human being, a completely arbitrary number.

Happens every day. NHTSA rates the value at saving a human life (VSL) at $9.1M. Insurance adjusters do it. Courtrooms do it. Etc...

And while arbitrary, it's not completely arbitrary. A lot goes into the figuring. Expected lifetime earnings, funeral costs, medical expenses, school costs, all goes into the figures.

Predicting extra illnesses relies on collecting data from medical studies subject to massive publication bias and never designed for that purpose.

Again: Educated guess. We know mercury in the environment is bad. Ergo, we set the fees for emitting it extremely high. $10M/ton might actually be an OOM low. $100M/ton might be closer.

Whatever. The idea is that you have a serious board that 'does their best', and regularly reviews the pollution list, adding new pollutants and adjusting the fees as new science gives you 'settled' values. I say settled because values shouldn't be assigned on the basis of 'cutting edge' science, but science that has been verified for a bit. If nothing else, it allows industry to keep an eye on the studies and have a good idea what they're going to be charged in the future.

As for parking/speeding tickets, first you'd have to define a goal. When your publicly stated goal is prevention, but your actual goal is revenue generation, of course the fee structure is going to look odd.

That is why people are getting upset with the regulatory overreach we are experiencing now. People are faced with arbitrary fines completely out of proportion to any harm they could conceivably be causing.

Ah, good thing I'm talking about fees, not fines. It's the difference between a sales tax and a speeding ticket. I also explicitly stated that the schedules would be set in proportion to harm being caused, at least as best as we're able to. Of course, that brings up a different point - it should probably charge different amounts depending on whether the pollution is affecting the ground, air, or water.

As for civil asset forfeiture, I laugh because I just got a nice letter back from my senator after I wrote him to ask for actions to seriously limit said seizures. I haven't heard back from the other congressional members(state & federal) that I also wrote, but oh well.

So I'll repeat: These are fees, not fines, in that they're not punitive. Things do tend to get crazy when you go punitive. This is 'simply' an attempt to render an external cost not paid by the business doing the industry, an internal one, so they have incentive to reduce their pollution. While actually REDUCING the regulatory burden, because now rather than sticking their noses in 'state of technology' and saying you have to have XYZ technologies installed, they simply charge for your emissions, and the business installs XYZ simply in order to save money. They even have incentive to seek new pollution control technology because it'll save them money, as opposed to the current situation where they don't want to see new technology because the EPA will subsequently mandate it.

Consider that Local Motors themselves said the cars are not street legal.

I had not considered the 'we're disruptive so laws don't apply to us' aspect. Assuming they want to operate as a legitimate company and not have a bunch of dead customers, I am sure even you could find some laws in here that would make a 'design it yourself' car made out of printed plastic just a bit of a problem, especially in the 'crashworthiness' section.

http://www.nhtsa.gov/cars/rule...

Alcohol makes you more aggressively confident, pot makes you more careful of your driving.

Not as such.

Acute cannabis consumption and motor vehicle collision risk: systematic review of observational studies and meta-analysis

Acute cannabis consumption nearly doubles the risk of a collision resulting in serious injury or death

Effects on Driving:

... Marijuana has been shown to impair performance on driving simulator tasks and on open and closed driving courses for up to approximately 3 hours. Decreased car handling performance, increased reaction times, impaired time and distance estimation, inability to maintain headway, lateral travel, subjective sleepiness, motor incoordination, and impaired sustained vigilance have all been reported. Some drivers may actually be able to improve performance for brief periods by overcompensating for self-perceived impairment. The greater the demands placed on the driver, however, the more critical the likely impairment. Marijuana may particularly impair monotonous and prolonged driving. Decision times to evaluate situations and determine appropriate responses increase.

A field sobriety test doesn't care what substance you've been imbibing. It tests your current level of impairment. Which is what we should be looking at if the goal is to reduce injuries and fatalities on the roads.

Why waste all kinds of money on tests that may or may not be able to measure actual impairment? And that goes for alcohol too.

Even the National Highway Traffic Administration says measured active THC levels can't be correlated with impairment:

They quote research that indicates driving while intoxicated by marijuana increases risk.

Effects on Driving:

The drug manufacturer suggests that patients receiving treatment with Marinol® should be specifically warned not to drive until it is established that they are able to tolerate the drug and perform such tasks safely. Epidemiology data from road traffic arrests and fatalities indicate that after alcohol, marijuana is the most frequently detected psychoactive substance among driving populations. Marijuana has been shown to impair performance on driving simulator tasks and on open and closed driving courses for up to approximately 3 hours. Decreased car handling performance, increased reaction times, impaired time and distance estimation, inability to maintain headway, lateral travel, subjective sleepiness, motor incoordination, and impaired sustained vigilance have all been reported. Some drivers may actually be able to improve performance for brief periods by overcompensating for self-perceived impairment. The greater the demands placed on the driver, however, the more critical the likely impairment. Marijuana may particularly impair monotonous and prolonged driving. Decision times to evaluate situations and determine appropriate responses increase. Mixing alcohol and marijuana may dramatically produce effects greater than either drug on its own.

That seems consistent with emerging research.

Acute cannabis consumption and motor vehicle collision risk: systematic review of observational studies and meta-analysis

Acute cannabis consumption nearly doubles the risk of a collision resulting in serious injury or death; this increase was most evident for studies of high quality, case-control studies, and studies of fatal collisions

Would it surprise you to learn that not all the data out there supports the position of the primary advocacy group for the legalization of marijuana?

Acute cannabis consumption and motor vehicle collision risk: systematic review of observational studies and meta-analysis

Acute cannabis consumption nearly doubles the risk of a collision resulting in serious injury or death; this increase was most evident for studies of high quality, case-control studies, and studies of fatal collisions

Effects on Driving:

The drug manufacturer suggests that patients receiving treatment with Marinol® should be specifically warned not to drive until it is established that they are able to tolerate the drug and perform such tasks safely. Epidemiology data from road traffic arrests and fatalities indicate that after alcohol, marijuana is the most frequently detected psychoactive substance among driving populations. Marijuana has been shown to impair performance on driving simulator tasks and on open and closed driving courses for up to approximately 3 hours. Decreased car handling performance, increased reaction times, impaired time and distance estimation, inability to maintain headway, lateral travel, subjective sleepiness, motor incoordination, and impaired sustained vigilance have all been reported. Some drivers may actually be able to improve performance for brief periods by overcompensating for self-perceived impairment. The greater the demands placed on the driver, however, the more critical the likely impairment. Marijuana may particularly impair monotonous and prolonged driving. Decision times to evaluate situations and determine appropriate responses increase. Mixing alcohol and marijuana may dramatically produce effects greater than either drug on its own.

Marijuana Might Make You a Worse Driver Than Alcohol Does

They described the five categories of vehicle automation, and explained that the first autonomous (not Musk’s so called “autopilot” which isn’t) vehicles will hit the road in the summer of 2015.

Here's the levels. Most high-functioning systems on the market, like the Tesla version, are in the Level 1-2 range.

No-Automation (Level 0): The driver is in complete and sole control of the primary vehicle controls – brake, steering, throttle, and motive power – at all times.

Function-specific Automation (Level 1): Automation at this level involves one or more specific control functions. Examples include electronic stability control or pre-charged brakes, where the vehicle automatically assists with braking to enable the driver to regain control of the vehicle or stop faster than possible by acting alone.

Combined Function Automation (Level 2): This level involves automation of at least two primary control functions designed to work in unison to relieve the driver of control of those functions. An example of combined functions enabling a Level 2 system is adaptive cruise control in combination with lane centering.

Limited Self-Driving Automation (Level 3): Vehicles at this level of automation enable the driver to cede full control of all safety-critical functions under certain traffic or environmental conditions and in those conditions to rely heavily on the vehicle to monitor for changes in those conditions requiring transition back to driver control. The driver is expected to be available for occasional control, but with sufficiently comfortable transition time. The Google car is an example of limited self-driving automation.

Full Self-Driving Automation (Level 4): The vehicle is designed to perform all safety-critical driving functions and monitor roadway conditions for an entire trip. Such a design anticipates that the driver will provide destination or navigation input, but is not expected to be available for control at any time during the trip. This includes both occupied and unoccupied vehicles.

U.S. Department of Transportation Releases Policy on Automated Vehicle Development

Here's a more technical discussion from NHTSA. At page 74-75, the data elements of the Basic Safety Message I and II are listed. The BSM Part I message doesn't contain the vehicle ID, but it does contain latitude and longitude. The BSM Part II message has the vehicle's VIN. So this is explicitly not anonymous.

Back in the 1980s, when Caltrans was working on something similar, they used a random ID which was generated each time the ignition was switched on. That's all that's needed for safety purposes. This system has a totally unnecessary tracking feature.

Most of this stuff only works if all vehicles are equipped. It also relies heavily on very accurate GPS positions. However, there's no new sensing - no vehicle radar or LIDAR. The head of Google's autonomous car program is on record as being against V2V systems, because they don't provide reliable data for automatic driving and have the wrong sensors.

If something is going to be required, it should be "smart cruise" anti-collision radar. That's already on many high-end cars and has a good track record. It's really good at eliminating rear-end collisions, and starts braking earlier in other situations such as a car coming out of a cross street. Mercedes did a study once that showed that about half of all collisions are eliminated if braking starts 500ms earlier.

V2V communications should be an extension of vehicle radar. It's possible to send data from one radar to another. Identify-Friend-Foe systems do that, as does TCAS for aircraft. The useful data would be something like "Vehicle N to vehicle M. I see you at range 120m, closing rate 5m/sec, bearing 110 relative. No collision predicted". A reply would be "Vehicle M to vehicle N. I see you at range 120m, closing rate 5m/sec, bearing 205 relative. No collision predicted". That sort of info doesn't involve tracking; it's just what's needed to know what the other cars are doing. It's also independent of GPS. Useful additional info would be "This vehicle is a bus/delivery truck, is stopped, and will probably be moving in 5 seconds.", telling you that the big vehicle ahead is about to move and you don't need to change lanes to go around it.

It's all quite useful but the real question is, is there that many deaths / accidents because of drowsiness?

The 1996 report from NHSTA says 56,000 in the US annually and more recent information indicates 110,000 incidents annually, though the injury and death rates remain the same between both claims. Both are also considered under-reported.