Nissan Develops Emergency Auto-Steering System

AmiMoJo writes "Japanese automaker Nissan Motor says it has developed a new technology to help drivers avoid collisions. A new computer system automatically steers the car to avoid colliding with objects in the road. The system relies on radar and laser scanners. It also uses a front-mounted camera to provide information on what's happening outside the car. The system first alerts the driver to turn in a certain direction. If the driver cannot immediately turn in that direction, the system takes over the steering to help avoid a collision."

Comments

recipie for disaster

[Anonymous Coward]

How well does said auto-steering system perform on ice, mud, or fording small streams?

Re:recipie for disaster

[LinuxIsGarbage]

Electronic Stability Control already does it on those surfaces.

Re:recipie for disaster

[hcs_$reboot]

The real problem is false positives: the car detects falsely a problem, avoids a non-collision, and even brakes by mistake. Worse will be when a false positive will induce an accident that wouldn't have happened otherwise.

Re:recipie for disaster

[Nexion]

Aye, learned that the hard way the first time my RX8 met ice. Pro tip... turn off traction control prior to hitting ice. Your speed will not matter. It would send me into a spin at 20mph.

Re:recipie for disaster

[Gaygirlie]

I've actually met several ambulance drivers around here who say the same thing -- these days they simply outright refuse to even consider cars where you can't turn such crap off. Those things are just a major disaster waiting to happen.

Re:recipie for disaster

[DogDude]

Electronic stability control is dangerous and a bad idea.

Re:recipie for disaster

[adolf]

ASC+T works great on my E36 BMW with some slight modification, though it wasn't at all bad in stock form.

Pro tip... install appropriate tires prior to driving on ice. I've got a set of skinny Blizzaks that I use for winter driving, and the combination makes the E36 the most stable and predictable car I've ever driven on snow and ice. It just works.

Re:recipie for disaster

[dubbreak]

Pro tip... install appropriate tires prior to driving on ice.

This.

Proper tires make all the difference. I have a FWD Rav4. Stock "all season" tires would cause it to go into a traction control seizure on slippery inclines (it would just shudder until you turn the traction control off). With some proper winter tires (full studable winter, not "winter rated") it was great in the snow and ice. I tried to get it out of control on purpose and between ETC and ESC it would right itself every time (this was in northern Canada with plenty of snow and ice in -20C).

Still have to watch for breaking though. If you are carrying too much speed and hit ice antilock isn't going to save you. Driving slow and engine braking will.

Re:recipie for disaster

[Jane Q. Public]

" I've got a set of skinny Blizzaks that I use for winter driving, and the combination makes the E36 the most stable and predictable car I've ever driven on snow and ice."

I live in an area that often has very snowy winters, and trying to convince people to use narrow tires in the winter is like talking to a brick wall. They want to put monster tires on their trucks, for example, and try to claim that the big "contact patch" will solve all their problems.

But I have seen the difference with my own eyes. Listen to parent, folks. If you drive on snow & ice, get narrow tires. It might just save your butt.

Re:recipie for disaster

[drinkypoo]

Electronic stability control is hard to do right but is now mandatory. It's still a great idea, but you have to be careful who you get it from, and yet everyone buying a new car is getting it.

Re:recipie for disaster

[drinkypoo]

Still have to watch for breaking though. If you are carrying too much speed and hit ice antilock isn't going to save you. Driving slow and engine braking will.

Engine braking can only be decreased by nudging the accelerator pedal, while normal braking can be varied. Fancier ABS than you get with a Toyota knows what to do in snow or on ice. And even with old ABS, you still get to steer, though you may not get to slow down appreciably.

Re:recipie for disaster

[Mspangler]

"Worse will be when a false positive will induce an accident"

That was my first thought too. Car sees a monster tumbleweed coming and swerves me into the other lane (boom) or the ditch trying to dodge it, not realizing that Ramming Speed is fully authorized with tumbleweeds.

Even worse would be the car dodging a big dog and hitting a small kid instead.

Re:recipie for disaster

[undefinedreference]

+1 on Blizzaks

I loved my Blizzaks when I lived in the snow country. I ran normal tires for years with good results (after all, where I lived you only had 5-10 days with snow on the ground to worry about in a normal year) until the year I did a project in Wyoming and decided it was a good idea to get snow tires. These tires are unbelievable, even on straight ice. They looked really weird, but their performance made me a believer.

Re:recipie for disaster

[dubbreak]

I didn't mean dropping it a gear in an emergency but rather driving in a lower gear at a slow speed so when you lift the accelerator you have the nice gradual braking. Decelerating in that way guarantees you have rolling friction rather than static friction of locking wheels. That's the aim of ABS, to have rolling friction rather than static locked wheels. Regardless of how fancy your ABS is, driving speed is what's going to make the biggest difference in braking distance.

Re:recipie for disaster

[ThatsMyNick]

and it is also mandatory.

Re:recipie for disaster

[Anonymous Coward]

I've actually met several ambulance drivers around here who say the same thing -- these days they simply outright refuse to even consider cars where you can't turn such crap off. Those things are just a major disaster waiting to happen.

Hm. Depends, really. If they "fail" and cause fewer deaths or injuries than they save, isn't it a net gain?

Seatbelts have killed people. But they're still a net gain. Isn't that what matters? Or do we demand that safety mechanisms *never* cause problems? If so, their design would require that the benefits be considerably diluted.

Re:recipie for disaster

[ThatsMyNick]

Ambulance drivers are usually far better drivers than the rest. They also understand their vehicle and its response far better than the average folk. So yeah, no wonder they prefer to take decisions themselves than trust the machine to do that for them.

Re:recipie for disaster

[green1]

The anti-lock brakes on my work truck do this all the time, 1 tire hits a small patch of gravel/ice/manhole cover and suddenly I feel a drastic loss of braking power on all 4 wheels. It has nearly caused many collisions.
For some reason the anti-lock brakes on my personal vehicle, despite being very annoying, have never caused me the same problems, despite being on a vehicle nearly 15 years older (so theoretically not as advanced a design)
If I could find a vehicle without them I would definitely have it, for safety reasons!

Re:recipie for disaster

[danomac]

I was in a vehicle in which the ABS system decided to malfunction. There's nothing like the feeling of standing on a firm brake pedal and noticing you're not slowing down. Thankfully is was at 10 kph. I was also glad there wasn't a pedestrian crossing the street at the time.

New technologies do have some issues. Although I really wish that system would have failed off than failing on (thinking I was in a skid.)

Re:recipie for disaster

[ThatsMyNick]

Did the emergency (or the parking) brake fail too?

Re:recipie for disaster

[LordLimecat]

I am curious as to why engine breaking vs normal breaking would make a difference. As I understand it, slipping happens because of inertia, when you decelerate too quickly. I cannot think of a reason why it matters WHAT is robbing your car of speed, whether it is at your discs or in the engine or because you turned too quickly.

If Im mistaken here, I would love to understand why.

Re:recipie for disaster

[LordLimecat]

parking brakes are supposed to be fairly terrible at stopping a car in motion.

Re:recipie for disaster

[v1]

easy fix. open the fuse box under the hood. look for the one labeled "ABS". Pull. done.

Re:recipie for disaster

[Russ1642]

RTFM. The manual will tell you to turn off traction control on snow and ice. I live in Edmonton, and I can't get my car down the street in the winter unless I turn off traction control. As soon as one tire slips a tiny bit it'll cut power so you can't get your speed up. If you're in deep snow it's a sure way to get stuck because you have to be able to spin your tires a little bit as you drive.

Re:recipie for disaster

[ShadowRangerRIT]

I've run into this on my car. I've got radar based crash avoidance (it's just brakes, no steering assist); it sometimes detects an imminent collision for a fraction of a second just before crossing railroad tracks. Luckily, it's so quick that I get the audible alert, but the brakes don't kick in. It's disconcerting though. If it took steering control, that would be terrifying.

Re:recipie for disaster

[ThatsMyNick]

It is decent if you dont lock your wheels. It may take twice as much distance to stop compare to your regular brakes, but are really good as a backup.

Re:recipie for disaster

[hairyfeet]

Have a buddy who was a county cop and said the same, he wouldn't let any of his family anywhere near a vehicle with the electronic traction control as he said 9 times out of 10 it would kick in at exactly the wrong time and cause an accident instead of avoid one.

I think the problem is they test this stuff on tracks where everything is rigidly controlled when IRL we all know its nothing like that, ice can be patchy, same with slick spots and gravel slung across the road, and these things have an "all or nothing" approach that goes too far in compensating. I know I wouldn't want my family in one, not after hearing some of my friend's stories on accidents he's worked where that crap was involved.

Re:recipie for disaster

[similar_name]

My understanding is that static friction is greater than kinetic friction. For instance when you push say a sofa across the floor a greater force is needed to move it from rest (static friction) than to keep it sliding (kinetic friction). If you're tires are rolling they essentially keep static friction because the surface of the tire is static relative to the ground at any given point. Once the tires lock, kinetic friction takes over and your friction coefficient goes down and the tires slide. I believe engine braking would keep the tires rolling somewhat to maintain some level of static friction. ABS does the same by 'pumping' the brakes quickly over and over in an effort to maintain as much static versus kinetic friction as possible.

While I believe I grasp the idea, my terminology may be off so someone else may be able to provide a clearer answer.

Re:recipie for disaster

[Jstlook]

Seriously! If I wanted my vehicle to drive for me, I'd take the bloody bus!

If they're so hot on not letting me actually drive a vehicle I own, why should I pay for it?

Re:recipie for disaster

[SimplyGeek]

You must be a member of PETA.

Re:recipie for disaster

[Quietust]

My guess would be that in low-traction situations, even light pressure on the brakes can be enough to lock the wheels, while engine braking cannot possibly cause the wheels to lock.

Re:recipie for disaster

[eggfoolr]

Engine braking is only on the driven wheels and you have less control of modulation. So for a front drive car it may induce understeer and loss of steering control and in a rear drive car it may induce oversteer (for experienced drivers this is preferable). Under most normal conditions engine braking is fine though.

On a 4x4 the engine braking is more even but may still give you mixed results compared to the cars braking system that is better balanced and with all the electronic assistance of ABS EBC ASC etc.

When in doubt put it in neutral and use brakes carefully. Again that's not necessarily the best result. A skilled driver can still modulate the engine revs to match road speed and gain more control. But 98% of people on the road are not skilled. (did I type that out loud?)

Re:recipie for disaster

[ThatsMyNick]

You buses drive themselves? Awesome! Which city are you from?

Re:recipie for disaster

[Jane Q. Public]

This.

When your tires are rolling, the friction in the direction they are rolling is far from zero, but it is vastly less (in normal road conditions) than from side-to-side. The "contact patch" of your tire is (relatively speaking) in a static position on the roadway, and so is under static friction. That is what keeps you going in the direction you point it.

When a tire is locked up, however, it is in dynamic friction. Dynamic friction is not only much lower than static friction, but since the tire is locked up, it no longer "cares" what direction it goes in, because the friction is exactly the same no matter what that direction is. Sliding is sliding, and the contact area is constant.

So if you are in a situation of rear-wheel lockup, but your front wheels are still rolling but braking, the rear of your car is going to "want" to overtake the front, and there is nothing at all keeping the rear from swinging out sideways and doing so, if there is any deviation at all from a perfect straight line. Not A Good Thing.

You can do this experiment with a small model car. Lock up the front wheels but not the back, and give it a good shove forward on a relatively smooth surface. Chances are, it will keep going more-or-less forward. Then free the front wheels and lock the back, and give it the same kind of shove. It will spin all over the place.

That is why for many years (before ABS), manufacturers would put disk brakes on the front but old-style drum brakes on the back. Because IF you are going to get a lockup, you want it to be your front wheels, not the rear.

Re:recipie for disaster

[causality]

My guess would be that in low-traction situations, even light pressure on the brakes can be enough to lock the wheels, while engine braking cannot possibly cause the wheels to lock.

When I needed a replacement car, I got a very good one at an even better price and have not regretted this decision ... but I certainly do miss a manual transmission for just this reason.

A secondary reason is that it's just more fun.

Re:recipie for disaster

[swalve]

Ambulance drivers are usually far better drivers than the rest. .

Not based on what I've seen.

Re:recipie for disaster

[Leroy Brown]

Hah.. I was along for a test drive of a new car once with a friend. We drove about two miles without him realizing the parking brake was full on. Upon noticing an odd smell eminating from the car, the salesman remarked that the car was so new that you could still smell the "engine surficant."

Re:recipie for disaster

[russotto]

I am curious as to why engine breaking vs normal breaking would make a difference.

Engine braking can't lock your brakes. If you try to use too much engine braking, your wheels will start to slip but not as severely as if you apply too much normal braking and lock them up. Engine braking also typically applies only to two wheels, which can be an advantage (sometimes, with RWD, so you can steer with the fronts while slowing with the rear) or disadvantage (most of the time).

I don't see why a good ABS wouldn't do better than engine braking -- it would maintain a fairly constant amount of slip. IIRC, good ABS does beat professional drivers on ice.

Re:recipie for disaster

[TheLink]

It has to be much safer. Maybe as safe as elevators. Because such stuff reduces the effect of your genetics in avoiding fatalities. ;)

For example say an automated vehicle is 10% safer. If people drove themselves there would be 10% more fatalities, but over a few generations it could get fewer and fewer (if the selection pressure is enough - if it isn't then traffic safety isn't that important compared to other pressures right?).

Similarly for the extreme sports stuff - it might be bad for some of the participants but good for the gene pool quality and spread ;).

Re:recipie for disaster

[ThatsMyNick]

Even if you dont agree that they are better, I am sure you would agree that, they think they are better drivers.

Re:recipie for disaster

[TheLink]

Or swerving off a cliff instead of hitting the side of it (perhaps intentionally - to slow down brakes fail etc).

Re:recipie for disaster

[houstonbofh]

Perhaps, and perhaps not. But they do have a larger sample of the failures...

Re:recipie for disaster

[swalve]

In the world maybe, but in the US, there are slightly more dogs than people under the age of 18 (78 million versus 72 million).

Re:recipie for disaster

[houstonbofh]

Or, will it see the empty shopping bag blowing down the road and swerve into oncoming traffic just out of range of the sensor? Or better yet, swerve off a cliff? And Toyota thought the carpet under the gas peddle was bad...

Re:recipie for disaster

[ThatsMyNick]

Mmm strange. Some states even have laws for these. Delware requires a vehicle traveling at 20 mph to be able to stop with 55 feet (and for regular brakes, it should stop in 30 feet). Have they stopped enforcing these laws?

Re:recipie for disaster

[jenningsthecat]

I think I can explain this. The first thing to understand is that the amount of force needed to make a tire 'break free' and skid, as opposed to continuing to roll, is less than that needed to *maintain* the skid. (This is analogous to pushing a heavy box along the floor - it's harder to get it moving than to keep it moving). When using your brakes, it's very easy to cross that 'breakout' threshold and put the car into a skid. And on slippery surfaces, the transition between rolling and sliding, (i.e. between the wheel rolling and the brakes locking up), is very abrupt.

But with engine braking, the wheels are always turning, by definition. The wheels can't lock, so they're less likely to reach the breakout threshold. This is very important for steering - if the wheels aren't rolling, the car basically won't respond to the steering wheel - the front tires become rubber plows. (FWIW, I drive an automatic, and I still downshift on slippery surfaces when I want to slow down - engine breaking isn't limited to standard transmissions).

While I'm at it, I'm in the mood to explain ABS. ABS applies and releases the brakes really quickly when the controller senses that the wheels are no longer rotating. Steering happens during the short intervals when the brakes are released and the wheels are rolling; slowing down/stopping occurs mostly during the transition from from rolling to locked, and partially during the interval when the wheels are locked. Keeping the wheels in that 'break free' zone as much as possible is not only responsible for maintaining steering ability in a skid - it also decreases the stopping distance. (This goes back to what I said in the beginning about the force needed for the wheels to break free being greater than the force needed to maintain the skid). BTW, in an ABS stop you have to turn the steering wheel a lot farther than you're used to for a given degree of arc, because the wheels are only redirecting the car roughly half the time - the rest of the time they're just sliding.

And just a reminder that if you're driving on a REALLY slippery surface, technology and mad skillz won't save you - you're at the mercy of inertia and external forces. I've been there, and it sucks.

Re:recipie for disaster

[danomac]

That's nothing. I still have the habit of having one hand at 12 o'clock on the steering wheel. I just got a car that has airbags two years ago.

  I really need to break that habit. I've been driving manual transmission cars for 20 years, it's rather hard to change something you be been doing for that long...

Re:recipie for disaster

[artor3]

No, electronic stability control is a great idea that makes cars much safer ...for most drivers.

For highly skilled drivers, it probably gets in the way more than it helps. But the vast majority of drivers aren't pros. And there's no denying that accidents have gone down markedly since ESC was introduced. And lest you claim that's a coincidence, studies have also shown strong correlation between vehicles that lack ESC and vehicles that end up in fatal accidents.

Seriously, you shouldn't make claims like that before looking at the evidence. If someone sees your post and decides to avoid getting ESC in their next car, their risk of death goes up by quite a bit. It's akin to telling people vaccines cause autism. Stop it.

Re:recipie for disaster

[Gaygirlie]

Hm. Depends, really. If they "fail" and cause fewer deaths or injuries than they save, isn't it a net gain?

Aye, I don't disagree. I just haven't seen any conclusive proof of such -- at most I've seen some "studies" from the manufacturers themselves, but no trustworthy studies from 3rd parties.

Seatbelts have killed people. But they're still a net gain. Isn't that what matters?

In a way, yes, but seat-belts can only hurt their wearers. A malfunctioning traction control system, malfunctioning brakes, malfunctioning steering -- they all have the capability of hurting a whole lot more people than just the passengers and therefore they should meet much, much higher standards IMHO. Collateral damage is always a Bad Thing(TM).

Re:recipie for disaster

[iinlane]

It's not very intuitive, but the contact patch is the same with narrow and wide tires. If the weight of the car and tire pressure remains the same, then also the contact area must be the same. Force equals area tmes pressure.

Re:recipie for disaster

[adolf]

But the shape of the contact patch is not the same. This is important because in the real world we use tires for driving, and not for static measurements.

Wider tires must push more material out of the way in order to find good purchase on squishy road surfaces, and they'll be able to utilize that purchase for a smaller percentage of a revolution than a skinny tire.

Re:recipie for disaster

[gman003]

Are you sure about that?

Factor in thinner tires having less air in them (as in, fewer molecules, or the volume of the gas in them at STP). You can take this to an absurdity if you wish - you could have a massive, building-sized tank of N2 at the same pressure as a tire, but the area is obviously different.

I'm not going to argue whether they work better on ice - I'm not exactly living in a cold-winter area, they close the schools if there's an inch of snow on the ground. I have zero capacity to judge whether they work or not. But the reasoning doesn't seem right. I would hazard a guess that it has more to do with the weight of the car being the same, but the contact patch being smaller, leading to greater force on the tires.

Re:recipie for disaster

[jamesh]

The real problem is false positives: the car detects falsely a problem, avoids a non-collision, and even brakes by mistake. Worse will be when a false positive will induce an accident that wouldn't have happened otherwise.

Such an anti-tailgating device already exists...

Re:recipie for disaster

[jamesh]

"Worse will be when a false positive will induce an accident"

That was my first thought too. Car sees a monster tumbleweed coming and swerves me into the other lane (boom) or the ditch trying to dodge it, not realizing that Ramming Speed is fully authorized with tumbleweeds.

Even worse would be the car dodging a big dog and hitting a small kid instead.

This is why I love slashdot. Full of people brilliant enough to imagine the possibilities of how something could go wrong, but too dim to imagine that the designers have already considered it :)

Re:recipie for disaster

[Jane Q. Public]

"It's not very intuitive, but the contact patch is the same with narrow and wide tires. If the weight of the car and tire pressure remains the same, then also the contact area must be the same."

This is not necessarily so. You are also assuming similar construction.

It depends a lot on the rigidity of the tire, and tires made in this fashion are typically more rigid than your standard street tire. So the same pressure will typically have a smaller contact patch.

Re:recipie for disaster

[adolf]

My work truck did this sometimes, too. The system would experience a tiny amount of wheel slip (which really does happen on gravel/ice/manhole covers, since neither tires nor suspensions are perfect), and overreact as if it had detected a large amount of wheel slip.

It would then modulate the brake for me, even though I rightly knew that it shouldn't have been doing this. And in doing so, it must modulate both front* brakes to some extent in order to prevent brake steer and the resultant unintended change of direction.

The problem was an intermittent front wheel sensor. It was replaced, with no other work performed at the same time, and things have been fine since then.

Time to grab a meter and/or a scope, along with a hammer and figure out what's up with your ABS. It's broken, and you should probably fix it...

*: It may have been doing something with the rear brakes as well, but I didn't notice this if it was. They don't do a whole hell of a lot on my vehicle.

Re:recipie for disaster

[Snotnose]

Err, no. My daughter was an EMT for a few years (quit when she got preggers with the best twins in the world :). She went through a pretty intense schooling period of about a year. It was all about anatomy and saving lives. None of it was how to drive.

Re:recipie for disaster

[fisted]

Counter-question, smartass: How well does average joe perform on ice, mud, or fording small streams?

Re:recipie for disaster

[Ironhandx]

Anything that has any sort of significant down side, including seat belts, should be optional.

According to studies done by the RCMP in Canada before they made seat belts mandatory everywhere seat belts were only responsible for prevention of death in 30% of the accidents in which they were a factor.(I.E. the car getting crushed is the car getting crushed, seat belt didn't really matter there) In nearly all cases they caused injuries and in 5% of cases direct death could be attributed to the seat belt being worn.

Seat belts are awesome when there are roll cages installed in the vehicle.... how many Corollas and Honda civics do you see out there with roll cages installed? Additionally the best seat belt in those cases is a proper 4-point harness. Not the "quick clip" seat belt invented by a company that thought it would make it easier to market. It did, it also made the thing almost as useless as a lap belt.

The incidence of neck-related injuries has risen nearly 900% since seat belts were made mandatory. Deaths per reported accident has only dropped about 20% and most of that can be attributed to better crash design in the car chassis and stricter safety regulations that have been introduced that have nothing to do with a seat belt. Given the down side is significant, seat belts should be optional.

Air bags should not be optional, the amount of damage they cause(broken thumbs, occasional cracked ribs, the odd broken nose or arm) is vastly outweighed by the damage they prevent. Some people say the air bag and seat belt work in sync, they do not. Modern air bags deploy so fast its already halfway inflated by the time the locking mechanism in your seat belt kicks in. All the seat belt does is give your head time to snap forward and nail the airbag while its near the end of its inflation and travelling at some ridiculous speeds that are the peak of its inflation speed. Hello whiplash and neck problems for the rest of your life as a best case, hello lifetime paralysis or death as worst cases. The seat belt didn't save you, the air bag did.

The same goes for ABS as for seat belts. The idea is your stopping distance is shorter. In many coniditions it isn't, but that isn't stopping idiots from lobbying to get it made mandatory here. Ice, deep snow, hydroplaning, all of these conditions fuck ABS sensors to hell. There are two theories(And these are straight theories, they can only be MARGINALLY proven in a lab, but the guy that invented the shit apparently had a marketing genius on staff) about ABS.

One: It will allow more control of steering while allowing maximum braking. This is patently wrong. Someone who has the sense to steer out of a situation will have the sense to be able to ease off the brakes a little when they feel the car starting to fish on them, and easing back on the brakes rather than having a motor kick in that disengages them almost completely is much much preferred. Those that don't have the sense are going to death grip the steering wheel and the ABS is just going to get them to whatever they are about to hit even faster.

Two: It'll make you stop faster. With modern brakes on dry pavement even at highway speeds? ABS isn't even worth talking about. Stopping distance increased by 20%+ over a driver with a 30 minute lesson on proper braking. You now occupy the air space of the vehicle in front of you. The speeds at which ABS becomes necessary exist only on the German Autobahn and roads like it. At those speeds your reaction time is so slow(in relation to how much time it takes for your vehicle to suffer enough jarring braking to lose stability) for easing off the brakes that its better to have a sensor do it for you.

That being said, most people are stupid, and ABS is a good idea. Should it be a default option on vehicles? Yes. Should I still be able to purchase a vehicle without it? Very much YES.

Snow and ice? My truck was horrendous. Then I pulled the ABS relay and figured out how to disable the traction control. Now it handles like it should, like a d

Re:recipie for disaster

[cheater512]

Unfortunately modern medicine thwarts Darwin most of the time these days.

That plus idiots often kill innocent bystanders, sometimes not themselves.
That doesn't help Darwin at all. The innocent bystander could have a IQ of 150 and the idiot with a IQ of 70 kills them..

Re:recipie for disaster

[dgatwood]

I accidentally did that in a test drive once, too. Needless to say, I did not buy that particular model of vehicle....

Re:recipie for disaster

[adolf]

Engine braking on cars with open differentials can allow for one tire to spin backwards in instances where one driven wheel has an abundance of grip (dry pavement) and the other doesn't (ice).

And while I must confess that this is obviously very amusing and sometimes quite fun, I'm not sure that it's very good for safety.

(Planetary gearsets are neat.)

Re:recipie for disaster

[AvitarX]

As I delawarian that's been through I inspection about a dozen times, I can tell you they never check the parking break.

they do check the regular breaks, all four wheels must work, with no pull (slam on the breaks at 5 mph in the inspection garage on a special platform).

They also don't check the aim of your headlights, but there are laws about that too.

Re:recipie for disaster

[adolf]

And just a reminder that if you're driving on a REALLY slippery surface, technology and mad skillz won't save you - you're at the mercy of inertia and external forces. I've been there, and it sucks.

Indeed true, although mad skillz help some: While it may seem incongruous to some, I've found that it's possible to drive too slowly on steeply-banked curves: Gravity's a bitch.

I discovered this one icy night on a familiar curve near my house. Everything was covered in a pristine and glass-like layer of ice. I was driving very slowly and carefully, and in the middle of that curved the back end of the FWD car got sucked downhill toward the inside of the curve by gravity.

I kept the car out of the ditch with luck* alone, and all was well.

Subsequently, I began accelerating before that curve to allow some equilibrium between centripetal and gravitational forces, before gently slowing back down immediately afterward**.

This balanced the problem, and the technique has served me well on countless other steeply-banked curves in all manner of weather.

I've seen a lot of cars in that ditch. Mine won't be one of them.

*: This is what most people consider mad skillz: Surviving unscathed in what appears to be a chaotic situation.
**: This is closer to mad skillz: Using experience to predict enough of the chaos beforehand that it ceases to be chaotic and instead becomes mundane.

Re:recipie for disaster

[adolf]

But do thinner tires have less air in the real world?

Usually, wider tires have a lower sidewall profile, while skinnier tires have higher sidewalls. (This is a generalization based upon real cars and their available, and reasonable, wheel/tire combinations.)

(I'll let you sort out the math. I only mean to raise the question.)

Re:recipie for disaster

[complete loony]

Correct. ABS forces your breaks to release *after* they have already locked up. To regain control, your wheels must start spinning again to match the ground. But if you're driving on ice, that may not happen.

Re:recipie for disaster

[beelsebob]

Unfortunately modern medicine thwarts Darwin most of the time these days.

No it doesn't, it just changes the fitness function. The fitness function has been changed by all kinds of things (including the creatures being tested at the time) for all time.

Re:recipie for disaster

[Gordonjcp]

Seat belts are awesome when there are roll cages installed in the vehicle.... how many Corollas and Honda civics do you see out there with roll cages installed?

All of them. Anything built for the European market has to have pretty good rollover protection, rather better than the protection afforded by a couple of lengths of scrap scaff pole welded in by Crazy Pete at the tyre-fitting centre.

Re:recipie for disaster

[adolf]

9 out of 10?

I use traction control all winter long unless I'm out playing with the car on purpose in a closed environment, and even then I turn it on from time to time just to see how hard it is to piss off*. It hasn't caused an accident yet, let alone for 9 out of 10 times it activates due to limited-grip conditions.

What is this, "new math"? WTF are you smoking?

*: It's hard to piss off, and the maneuvers that cause it to be pissed are equally bad with it disabled: There isn't much that can correct for throttle-lift-induced oversteer other than steering appropriately and applying more throttle, and neither traction control nor lack of same does that without the driver physically pressing on the accelerator.

Re:recipie for disaster

[wonkey_monkey]

Which is what, exactly?

Re:recipie for disaster

[whydavid]

For what it's worth, the NHTSA (http://www-nrd.nhtsa.dot.gov/Pubs/811206.pdf) continues to find (every time they repeat the study) that seatbelts are saving thousands of lives a year in the US alone. I would love to know how you've arrived at the conclusion that safety gains in cars are solely attributable to design improvements. And, while 20% may seem modest, you are failing to account for the steady increase in driver speed over that same time period. As for ABS, I'm not familiar with the literature, but I would submit that insurance companies probably don't give discounts for cars with ABS (a common practice, at least in my area) because they fail to improve driver safety. Strictly speaking, that strongly suggests only that ABS reduces the total economic costs to insurers, but I think it is pretty safe to use that as a proxy for the severity/incidence of crashes. And seat belts don't require roll cages to be effective. Are you under the impression that every accident is a gruesome scene involving a semi-truck side-swiping a VW bug? What percentage of accidents do you really think involves the roof crushing the driver during a violent rollover? And in what percentage of those do you think the driver was better off without a seatbelt?

Re:recipie for disaster

[michelcolman]

There are many different types of Electronic Stability Control. Unfortunately, most car builders just take the cheapest and most basic one available. If I compare the ESC in my wife's Prius with the one in my Merc, it's a world of difference. When you turn into a street with the Prius (with lots of gas because there's traffic coming), as soon as the front wheels start spinning a bit, the ESC kicks in and goes "oh, my, understeer, that's mighty dangerous!" and slams on the brakes. What the f*** do I care about understeer, there's traffic coming! I don't care if the turn is slightly wider! My Merc, on the other hand, will happily allow even a little bit of oversteer before, very briefly, making a small correction using one of the brakes while continuing to accellerate. You really have to go overboard before it makes a drastic correction, and trust me, that means you needed it.

I just love driving my Merc in winter. With everything on, it makes you feel like you're the best driver in the world. Of course it's more fun with it off, if you want to do doughnuts and things like that, but I've never felt less safe with it on. And trust me, I'm no wuss on the road.

Re:recipie for disaster

[SuricouRaven]

If current conditions stagnate technologically and socially for an evolutionary significent amount of time, we'd see the 'idiocracy' scenario. The stupid people breed a lot more, so the average intelligence of the population would plumet. Not likely to happen in reality though, as the time scale for evolutionary chance is many orders of magnitude greater than that for more immediate forms of change.

Re:recipie for disaster

[quasius]

Where are you supposed to be positioning your hands now?

Re:recipie for disaster

[beelsebob]

Such a change would be self limiting, as the idiocracy would cause the (possibly slow through gradual degradation) destruction of the medical systems keeping them able to function in the world... Once again changing the fitness function.

Re:recipie for disaster

[dotancohen]

Seat belts are awesome when there are roll cages installed in the vehicle.... how many Corollas and Honda civics do you see out there with roll cages installed?

All of them. Anything built for the European market has to have pretty good rollover protection, rather better than the protection afforded by a couple of lengths of scrap scaff pole welded in by Crazy Pete at the tyre-fitting centre.

This topic needs elucidation. The chassis modification usually referred to as a roll cage is _not_ designed to provide rollover protection, at least not the way they've been made in the past few decades. The modifications are designed to stiffen the chassis to provide better high-torque-from-0 traction, prevent flex which could crack windshields, and provide supplemental side-, front-, and rear- impact protection.

You are right, Crazy Pete does not have computer simulations or real empirical data about the roll dynamics of the particular model car he is welding in. However, supplementary impact protection is a no-brainer, so long as Pete understands other safety mechanisms in the vehicle, for instance by not welding in a door bar that would actually puncture the occupant in a side-impact collision, or prevent the stock side-impact airbag from deploying properly.

I assume that the modifications are termed roll cages due to the historical fact that pre-1970s vehicles offered almost no roll protection, so _any_ cage no matter how ill-designed was better than nothing.

Re:recipie for disaster

[realityimpaired]

Where abs comes into play, which the gp seems to have missed, is panic breaking. A calm driver who knows what they are doing can stop faster without them, but a panicked driver who is thinking about the big thing he is about to hit will lock the brake

Besides, a good abs system won't trigger unless you lock the brakes. Only time I have triggered mine has been when I hit black ice while braking normally, and I have been glad of it then. Similarly, good traction control is only there when you need it, and can be turned off. The owners manual for my 2011 Subaru Impreza even lists when, for safety reasons, it should be turned off.

Re:recipie for disaster

[dotancohen]

It's not very intuitive, but the contact patch is the same with narrow and wide tires. If the weight of the car and tire pressure remains the same, then also the contact area must be the same. Force equals area tmes pressure.

You are assuming that the air pressure carries the entire weight of the vehicle. In fact, this is not true: the air pressure carries n% of the weight of the vehicle, and the rubber carries the remaining (100-n)% of the weight. For different purposes different proportions are desirable: for instance large earth-moving vehicles like n to be as large as possible to prevent rubber wear. I don't know the engineering considerations behind winter-tire design, though.

Re:recipie for disaster

[michelcolman]

Your argument about ABS and isurers is flawed because it is based on the assumption that insurers are rational, intelligent beings. They are not. They probably just go "hey, this car has an extra safety device, the manufacturer says it's great, so we'll give a discount for it".

That does not mean that I disagree with the conclusion, though: ABS is probably a good thing overall because most people don't know how to brake properly.

Re:recipie for disaster

[dotancohen]

There is no such thing as an emergency brake, only a parking brake. There is no point of failure in a modern vehicle that would reduce the hydraulic pressure in both braking systems (yes, modern vehicles have two independent hydraulic systems going to the calipers) such that the cable-operated rear braking would provide assistance.

The only common points of failure are the pedal assembly (designed fail-safe, by the way) and the master cylinder. Master cylinder failure would either reduce pressure in one hydraulic system (front or rear seal failure), or equalize between them (center seal failure).

If someone were to sabotage the hydraulic braking system or pedal assembly, then the parking brake would be helpful in stopping the vehicle. Of course, that assumes that the parking brake itself was also not sabotaged.

Re:recipie for disaster

[drinkypoo]

I didn't mean dropping it a gear in an emergency but rather driving in a lower gear at a slow speed so when you lift the accelerator you have the nice gradual braking. Decelerating in that way guarantees you have rolling friction rather than static friction of locking wheels.

What? Who told you that? That is simply not true at all. We use higher gears in snow and on ice to reduce torque at the wheels to reduce the chance of wheelspin. We do NOT repeat NOT use lower gears, that only increases the chance that you will break the wheels lose in either direction.

Regardless of how fancy your ABS is, driving speed is what's going to make the biggest difference in braking distance.

Well, no. Not really. Modern ABS is smart enough to lock up occasionally on snow to maintain a pile of snow in front of the wheel, and to pulse a little harder on ice as well. So actually, what makes the biggest difference is your technology. The newest cars will stop shorter from a higher speed than the older ones with ABS. And unfortunately, this doesn't include your average Toyota or what have you.

Re:recipie for disaster

[YttriumOxide]

Aye, I don't disagree. I just haven't seen any conclusive proof of such -- at most I've seen some "studies" from the manufacturers themselves, but no trustworthy studies from 3rd parties.

I don't disagree with you at all, but I'm leaning towards traction control being "in general" a good thing purely from personal experience.

I generally drive sports cars and that's what I'm used to. So, when being used to a "harder" feeling of the road and excellent handling at high speeds; I need to be significantly more careful when I get behind the boat of larger "family" or "luxury" cars (to me, most feel like piloting a boat rather than driving a car).

I had such a monster vehicle as a company car once. I wanted to see exactly what the traction control was doing for me, so I gave it a test on a wet road going around a small roundabout (after taking all necessary safety precautions regarding other vehicles). First time through, with traction control, at 40km/h; no problem. Second time, without traction control, 40km/h, same amount of steering; I spun around around 270 degrees before stopping.

Of course, even WITH the traction control I wouldn't normally take a tight roundabout on a wet road at 40km/h in such a vehicle. But, at least it gave me a very good feeling for what it does. Once I got used to it in that car, I knew exactly when it would kick in and what it felt like when doing so.

In my own sports car, I leave it on for normal city driving and am also now used to the feel of it and know when/how it behaves. Before having done those tests in my old company car though, I never turned it on. I only turn it off when racing or otherwise driving the car in a "sportier" manner.

All of that said, I'd also like to see some real trustworthy studies to back up my feeling and anecdote (or refute it). And also all of that said, I'm very uncomfortable with the idea of the car actually steering for me as described in TFA. I'd definitely want an even higher level of evidence for that than I would for traction control.

Re:recipie for disaster

[Anonymous Coward]

If current conditions stagnate technologically and socially for an evolutionary significent amount of time, we'd see the 'idiocracy' scenario. The stupid people breed a lot more, so the average intelligence of the population would plumet. Not likely to happen in reality though, as the time scale for evolutionary chance is many orders of magnitude greater than that for more immediate forms of change.

The effect isn't as profound and fast as people think. For example, if two geniuses mate, their child statistically will be far less intelligent than either of its parents, though still above average intelligence. And if two morons mate, their child statistically will be far more intelligent than either of its parents, though still below average intelligence. Right now, the greatest impact on intelligence isn't genetics--it is nutrition.

Want to know something scary? Think about how stupid the average person is. Two generations ago the average person would have had an IQ that equates to 80 today.

Re:recipie for disaster

[L4t3r4lu5]

This, and increase your braking distance. Riding up the guy in front's chuff in ice is asking for a big hike in your insurance premiums.

Re:recipie for disaster

[L4t3r4lu5]

Wide tyres offer a higher surface area to the snow surface, spreading the load, meaning it compress the snow under the tyre producing water. Water on ice is bad. Small tyres have a lower contact area, meaning they have a higher pressure on the snow, meaning they press deeper to the road surface.

Ultimately, though, if snow / ice driving is at all a regular thing for you, get some steel wheels and put studded snow tyres on, or invest in a set of snow chains.

Re:recipie for disaster

[arth1]

You must be a member of PETA.

Nope. I just don't believe that humans have an inherent special status or obviously should hold special privileges.

I'm not against animals killing each other for food. Whether it's my butcher killing a farm animal for feeding me or my neighbour's cat, or a puma killing and eating a kid (of either the four legged or two legged variety).
But would I kill a non-human animal to save a human? Nope.

A mother would kill the last remaining Vaquitas in the world to feed her starving child. I would rather feed the mother to the Vaquitas. That would serve a porpoise.

Re:recipie for disaster

[Skal Tura]

First time i drived a ABS enabled car on ice+snow with friction winter tires, not studded, was the last time i will drive a car with ABS + friction winter tires, even if they are brand new like in that case. Couldn't stop at all :(
The ABS wouldn't allow me to brake efficiently, it took all braking power away, finding the maximum amount of brake without ABS kicking in was the only way to even attempt stopping and dancing at the edge, any less brake and braking distances would have been at least 10 times longer than normal.
Thing is, most systems fail to recognize that slippery conditions such as snow or ice REQUIRES certain amount of slip for maximum grip, in snow it is as high as around 20%!

The only acceptable electronic traction control system is the Skyline ATTESA, as that system provides you with just enough oversteer that you will not end up in the ditch or tree: http://en.wikipedia.org/wiki/ATTESA

Then again i'm not joe average driver, i actually know a few things about driving.

Re:recipie for disaster

[ircmaxell]

The only common points of failure are the pedal assembly (designed fail-safe, by the way) and the master cylinder

And the ABS valve body assembly. Which I had go on my catastrophically on a 1994 Chevy Blazer. In that has, the only brakes I did have was the parking brake cable assembly.

The more complicated vehicles become, the more failure modes are possible...

Re:recipie for disaster

[Skal Tura]

Oh yeah forgot to mention BMW E90 traction control system is outright dangerous in conditions where you need to avoid hitting an obstacle.
Just at the wrong time it takes away all power, so you cannot return to your line safely, results most often in a spin.

It's a heavy car so after heavy braking while avoiding something, it does still feel good, it looses just enough grip on the rear to allow you some very good response into avoiding the collision, but when you start to flick it back (ie. not to wind up into a ditch) you are essentially doing inertia drift (feint drift for some of you), the problem is to pull it off you need some significant slip on rear tires for control, problem is that the BMW system pulls of all power just at the wrong time, either causing one of two things:
  - Rear tires grip and you get a pendulum inertia drift effect, you know after a weight shift if you do weight shift at correct time you get that much stronger weight transfer, causing you to spin out of control to the opposite direction you are trying to get.
  - You cannot apply sufficient force forwards causing at least half a spin to the direction you wanted to go, 90 to 180 degrees, with insufficient steering lock there's not much that can be done. Sounds counter intuitive i know "shouldn't it just straighten out then? You did a mistake earlier!" Nope, since tires have better grip forwards applying force allows you to also push the front and by weight transfer enables more grip on rear tires.

I liked that car A LOT, but driving the stability control on at slippery conditions was outright dangerous. Taking it off tho, BMW E90 series is VERY GOOD, it surprised me A LOT how can that heavy of a car be so nimble, agile and easy to control.
I haven't taken the car to public roads, just drifted it so don't know how good it is at roads, but knowing BMW it's probably very good.
E90 is one of the very few newer cars i'd like to actually own and use as a daily driver... As long as i can figure out a way to have the stability control off at all times and still have it insured :)

Re:recipie for disaster

[Skal Tura]

Except evolution has been speeding up in the past century - significantly.
That's because travel is so much easier and there is so much more diversity everywhere.
http://news.nationalgeographic.com/news/2007/12/071211-human-evolution.html

Further some scientists have argued that today evolution of the way we think is very rapid, can't find reference right now :(

100 years ago it was rather rare to see chinese men in america, or african at japan. Nevermind places like Finland where i live.
Still very little cultural mix in most Finnish cities, but in large cities there's a lot of variety. Just here in Helsinki there is A LOT of people form Africa, Turk, Philippines, China, Thailand, Vietnam, Czech Republic, Sweden, Germany, UK, Israel etc. From all everywhere in the world, especially a lot from Russia and Estonia as they are our neighbours.
I know quite a few people who's wifes/girlfriends are from outside finland.

Yes, there has been foreigners for centuries, but in modern day long distance travel is in the reach of all but the poorest!
I even know couple of people who has married with someone coming from very long away. One of my best friends married a mexican girl.

Now tell me that won't speed up things vastly! ;)
Yes it still takes generations to weed out the most desireable genes, but now more genetic variety is being introduced to the pool! :)

Re:recipie for disaster

[h5inz]

It seems to me that you are advocating the use of air bags without using seat belts. Here are some references:
http://www.ncbi.nlm.nih.gov/pubmed/18365327
http://www.ncbi.nlm.nih.gov/pubmed/12169939
Airbag helps to very much lessen the incidence of the neck-related injuries that are the biggest problem with seat belts. So I would rather say that air bags and seat belts are good in combination, which is a common belief as well.

Re:recipie for disaster

[BetterSense]

Where I grew up in the Midwest, (late 90s - early 00's), it was standard practice to yank out the ABS fuse from the fuse block during winter time. The rationale? Wanting the goddam brakes to work. Even ABS true believers eventually converted after a few episodes of, you know, the goddam brakes not working.

I understand that it's all different now, that modern vehicles have reached a state of technological advancement, that vehicle engineers design all-knowing algorithms that transcend physics itself. So I've heard.

Re:recipie for disaster

[Skal Tura]

I'm sorry but i got to say you don't have much experience driving cars hard, while better cars WILL make you faster, you should be able to take a corner of your description.

Learn counter steering, weight transfer and get a feel on the car, most modern cars, even sport cars, lack a proper feel. If you want a new car to learn with i suggest GT86 from Toyota, excellent response from the chassis i hear. Doubt it's anywhere near what the car's icon was, but closest thing to the real if you want a new car.

Not sure what's tight roundabout in your sense, and were you trying to go full 360 on it, or straight, or a 90 degree turn to the right, ie. the first out.
Now, go and do the same on snow (with STUDDED tires also containing friction ribbons or whatever they are called in english) and take it at 80km/h entry speed maintaining at least 60km/h exit speed if you are taking first or going straight.

Better yet, take a small, nimble, agile RWD car to a track day, you WILL have a blast while learning to drive better. Just start slow and at each lap go slightly faster. At the go too you could take it to a tiny autocross setup and learn dynamics of grip and weight transfer.

But you are absolutely right, regular cars have a boat feeling to them. I got gifted a Calibra as a daily driver, it's sufficiently old that it only has ABS and very low fuel consumption so i use it for daily driving etc. while my proper cars are at garage, it's absolutely terrifying experience at any speed above crawling when exiting corners, being automatic and FWD. Fortunately i'm going to use for the winter my '83 Corolla with a '97 1.6L 20valve VVT-i engine, roll cage, heavily modified suspension, heavily stiffened chassis with about 160whp and now that i finish this round of upgrades HOPEFULLY weighing in at just 840kg :)
The next model of that car is still the best in F-group rally. Shares same chassis bottom plate even, almost everything can be just bolted in from the next model to this older model, but this older model as Sedan model is some 150kg lighter from the factory.

Re:recipie for disaster

[L4t3r4lu5]

Pressure = force x area. If you still don't know why you're wrong, you should go retake grade school Science.

Re:recipie for disaster

[FatAlb3rt]

Minor nitpick - ABS reduces braking force when the wheel spin rate doesn't match the speed of the vehicle. The wheel is not going from 60 to 0 to 60 mph several times per second.

Re:recipie for disaster

[Ironhandx]

What you call a Roll Cage isn't one at all as they come from the factory, its chassis enhancements designed to compensate for the lack of a roll cage. I'm not sure where exactly you are, but a roll cage is a series of welded together tubular steel bars, generally around 2" in diameter with 1/4" thick steel making up the tube.

What they do to cars at the factory is give them a lick and a promise and send them on their way. Its fine if you roll once, at low-ish speeds, if not you're going to be wishing you'd added the extra few hundred pounds of steel that makes up a roll cage.

As an example of said substructures effectiveness: Take apart ANY 4x4 capable truck chassis made in the last 10 years. Its roll cages galore. The purpose of a roll cage is for the ROLL CAGE to survive impact from any side and protect the occupant by keeping him/her suspended in the middle via a harness system. The human body is incredibly resilient as long as it doesn't suffer any major lacerations or direct blunt force trauma so they are incredibly effective.

They're expensive to install as stock on a vehicle however so 95% of vehicles sold don't have one incorporated.

Also, we actually have stricter safety laws here, and anything but a low speed flip or a slide across a grassy field the additional roll protection does nothing, it fails with the first impact of any of the 6(particularly the 4 making up the exterior corners of the top of the cab space) supports against a rock because its entire structure is so dependant on the rest of the structure being intact. Which for the purposes of the safety test lab does just fine.

It's better than what there was 20 years ago(aka nearly nothing, your figure of pre-1970 is way off unless things were forced to be a lot different for euro travelers. They started the engineering process back then, they didn't actually get anything much out of it until they started getting away from all these box-shaped cars), and makes seat belts a bit more viable. The problem is they made the seat belt laws more than 20 years ago, and there are still situations where its safer to not be wearing one.

The trouble with the post-1970's up to about 96-98 cars was that there was some semblance of a roll cage but in many cases it would just make things worse as it was harder to remove after it inevitably pancaked right down to the doors during a roll.

There is an additional benefit that a roll cage will stiffen a chassis for better handling, but a straight bar across the back of the front seats accomplishes this task much more effectively and efficiently, so it would never be the primary reason for installing a roll cage.

Additionally you can purchase after-market roll cages with thick mandrel-bent tubing for most vehicles these days. While crazy-pete could screw something up, these folks make a business out of it, word would get around pretty fast if they made a shoddy product.

I've been in rollovers in vehicles with roll cages, and from the remains of vehicles without roll cages I've seen in rollover accidents along the highway I'm glad I've never been in one of those.

Worst thing that happened in a rollover so far is I fractured my right arm against the shifter.

Re:recipie for disaster

[Ironhandx]

Insurance companies here do not give out discounts for ABS equipped vehicles. They did for awhile, then they got new stats in and found out that ABS causes more accidents(and thereby $) than it saves. Now this is probably specific to area and I'm sure there are areas where ABS would perform better. If you have deep snow on any sort of regular basis for 2-3 months out of the year however it loses everything it gained.

Re:recipie for disaster

[somersault]

I would have thought constantly introducing more variety into the gene pool would slow down "evolution" - if by evolution you mean creating new species and traits. If we all keep interbreeding then wouldn't new traits take far longer to take hold? Everyone would be more similar, rather than having racial traits as we have just now. Not that I have a problem with any of that really :) But for cool things to evolve, generally they have to be isolated from the rest of the world - just think of all the interesting marsupials in Australia for example.

Re:recipie for disaster

[tibit]

You're silly. Traction control is inactive when you're not applying throttle. If you're hitting ice with throttle down and get the wheels to spin, well, traction control will try its best to limit engine power (and perhaps apply brakes to slow down spinning wheels on models with active brake pump modules).

Re:recipie for disaster

[Ironhandx]

Overall, yes, but as it stands right now they are actually making it harder to disable because it comes stock on everything. If its stock, there is no need to separate the system at all for easy disabling/removal.

Re:recipie for disaster

[BlackSnake112]

You have 4 point seat belts and a roll cage like those in race cars on regular cars in the EU as standard? No, you don't. The roll over protectio that you speak of is not a roll cage.

Crazy Pete might not be the one to go to to install a real roll cage into the car. I have seen a few cars (yes cars not trucks or SUVs) with a real roll cage added in. They did a lot more then weld in a few poles inside the car. The whole passenger compartment was reenforced inside, on the sides of it, and underneath the car. They took apart the doors and added in parts for reinforcement. That is far above standard for those cars. Most of them were older cars being used for racing. They still passed inspection and could be driven on the road legally. The point is the added in roll cage is far better then what come standard on cars.

Re:recipie for disaster

[tibit]

You're mixing up traction control with stability augmentation, unless you were really accelerating on that roundabout. Traction control limits engine torque applied to the wheels when said wheels lose traction. Stability augmentation uses individual control of brakes to provide steering torques that are not available from steering system alone. This is what most dolts out there fail to get, and I get really upset about it. On a classic car there is no lever to actuate to provide steering torques using the brakes -- yet applying such, based on your steering wheel inputs, is precisely what stability augmentation does! Asymmetric braking can provide you with very nice steering torques that can exceed whatever you get from turning the front wheels alone. Arguing that it's a bad thing to have this capability is going full retard. There are no asymmetric braking controls on stock cars without stability augmentation. With stability augmentation your steering wheel inputs, in certain conditions, are used to apply just such asymmetric braking. This works in concert with ABS sensors to ensure that wheels don't lock up and all available traction is used up for steering purposes. About the only thing you'd need in a sports car is aural indication of how much augmentation is applied. This would have let you figure out in real time how much margins you have left before you'll lose control.

Re:recipie for disaster

[fnj]

Second point is categorically incorrect. GP is right. Speed has by far the most dramatic effect on stopping distance. It's a SQUARE function. If a perfect driver or perfect ABS can stop in 140 feet from 60 mph, he or it can stop in 35 feet feet from 30 mph. ABS performance on the other hand is essentially a LINEAR function.

Yes, either driver aptitude or ABS performance have an effect, but it's not nearly as dramatic as speed.

Re:recipie for disaster

[YttriumOxide]

Thanks for all the info - I'm sure it will be invaluable to someone who isn't aware of most of that...

You do seem to have misinterpreted my post though. I was referring to deliberately driving the car in a way that I knew (or at least strongly suspected) would cause it to spin out without traction control but not with... i.e. actually testing that the traction control does what it's supposed to do. I certainly COULD have controlled the car to avoid spinning out, but that would've defeated the purpose of the test.

For reference, I make no claim to be an excellent driver; but I do have a general idea what I'm doing behind the wheel.

Re:recipie for disaster

[bondsbw]

If people drove themselves there would be 10% more fatalities, but over a few generations it could get fewer and fewer

As if computer engineers are going to sit around twiddling their thumbs for "a few generations".

By that time, computers will likely be so sophisticated that it would be considered a crime against humanity to drive a 2 ton vehicle manually.

Re:recipie for disaster

[dywolf]

The patch is -not- necessarily the same. that is an untrue statement.

You have to consider the strength dynamics of the tire shape overall. that is the the walls and structure of the tire begin affecting the contact area. tire width, tire height (road to rim), tire thickness, all affect the structural support properties of the tire. one of the properties of wider tires is they generally have a larger footprint. larger footprint plus same weight equals less bearing pressure, which is why larger tires are used when sinking into the surface is a concern (dirt, sand, mud). but not all wider tires; the "low profile" tires can be wider but not deform as much because the effective column height is much shorter. but it will still tend to deform in the middle ("flat" contact dimension) more than a narrow tire of same low-profile height would; think a beam type bending moment, but in reverse (upward). that upward force btw is why the proper inflation pressure is important and improves traction (and wire steel wire meshing/banding was added to tires years ago); more pressure to resist it, since the rubber itself has poor resistance to deformation.

Re:recipie for disaster

[Ironhandx]

I didn't say solely, I said in large part. Air bags and some semblance of an incorporated roll cage are certainly responsible for a much larger chunk of that than seat belts. Unfortunately the government only goes as far as the correlation and doesn't look into the causation most of the time.

Rollovers crushing the driver as a % of fatal accidents are fairly common(around 30% of fatalities I think? Have to look it up again), but these aren't even the accidents I'm talking about. The severe neck injuries and paralysis caused by the seat belts from plain old head on, and rear end collisions is more significant than the death, but when these things are regulated, any crash you survive, not necessarily one you can walk away from(even after a recovery period), is a win for them.

Side-impact of a VW bug into another VW bug can cause the same results as the semi into a VW bug as far as side-impact accidents go. Side-impacts are relatively rare(10% or less in most places, but I imagine this might increase in cities). The semi vs VW bug accident is actually much more likely to end up in a situation where whatever you have in the car or strapped around you short of a serious off-road full protection roll cage isn't going to make a difference. The roll cage only has a half a chance of making one.

However side-impacts are less safe now than they were years ago. The design of center consoles etc for creature comforts means your hips generally get crushed against it. Some cars don't have this problem but a lot do. The old vinyl-seat straight across with only a shifter in the way was better as far as side-impact goes. You take the blunt force, but you don't take the crushing as often, and the crushing is what will almost invariably kill someone. The blunt force is survivable.

Re:recipie for disaster

[YttriumOxide]

You're mixing up traction control with stability augmentation

You're completely right...

Anyone else: Please re-read my post with the words "stability augmentation" (or similar) anywhere it says "traction control" (and feel free to call me names for mixing it up).

Re:recipie for disaster

[drinkypoo]

ABS performance on the other hand is essentially a LINEAR function.

Who told you that? There's worlds of difference between ABS implementations, and that makes as much difference as anything. Old-school ABS effectively won't stop you on ice at any speed, and won't stop you in many snow conditions at any speed. It slows the car gradually, but not quickly enough to stop you for any meaningful term of the word. Again, it permits steering. But more modern ABS has new strategies that make it capable of functioning meaningfully in scenarios where the old stuff might as well just curl up and die.

Yes, either driver aptitude or ABS performance have an effect, but it's not nearly as dramatic as speed.

How much speed are we talking about? Because I'd rather have modern, snazzy ABS and be going 35 on an icy road than have antique ABS and be going 25.

Re:recipie for disaster

[tibit]

I think you have no clue how stability augmentation actually works. Stability augmentation not only provides additional steering torque from asymmetric braking, but will also control weight transfer between the wheels to maintain sufficient traction on all four wheels. It does it an order of magnitude faster than any human could do it in feedback control (i.e. without pretraining). With pretraining, a human is able to apply feedforward control -- i.e. brake inputs that are anticipatory in nature. Such feedforward control is only available periodically when you're paying full attention. I'm sure as heck drivers on a racetrack can be quite spent after just a few laps. Now imagine having to do an 8 hour stretch that way, ha ha. Alas, stability augmentation is always there, never tired, never bored, never distracted. Yet the control inputs available to the human are very crude: only one brake pedal. The stability augmentation system has individual control of all four brakes. With such understanding, perhaps you'd be able to get it into your head that a stability-augmented race car would not only make you safer, it'd give you very clear advantage over dolts who insist it's bad for them. The uninformed arguments against stability augmentation are eerily like from those, 80 years ago, who were against electric lights. In retrospect it makes you look very, very silly.

Re:recipie for disaster

[dywolf]

The world is full of engineering disasters that occured because the designer missed one little thing.
Such mistakes improve the world of engineering by learning from mistakes.
But peer review also helps.

Re:recipie for disaster

[tibit]

Some people say the air bag and seat belt work in sync, they do not. Modern air bags deploy so fast its already halfway inflated by the time the locking mechanism in your seat belt kicks in.

I don't know where you live and what you consider modern, but an otherwise pitiful '00 Volvo S40 has pyrotechnic seat belt tensionsers controlled by the SRS control module. That's a 12 year old model. That makes your statement 12 years behind the times, at least. The mechanical locking mechanism in the seat belt spool is pretty much there for use when braking without impact, and as a backup in case the SRS system is inoperative.

Re:recipie for disaster

[omnichad]

Modern air bags deploy so fast its already halfway inflated by the time the locking mechanism in your seat belt kicks in. All the seat belt does is give your head time to snap forward and nail the airbag while its near the end of its inflation and travelling at some ridiculous speeds that are the peak of its inflation speed. Hello whiplash and neck problems for the rest of your life as a best case, hello lifetime paralysis or death as worst cases. The seat belt didn't save you, the air bag did.

As someone who's been punched in the face by an airbag, I'll say that the best case is friction burns on the face and neck as that's all I suffered. I tend not to tense up my muscles suddenly so I'm not very prone to whiplash, and my neck suffered nothing at all other than a little stiffness for a few weeks. I'll admit I was the only one with injuries when hitting a large deer head-on. Everyone else's injuries were prevented by seatbelts.

Re:recipie for disaster

[Politburo]

You're talking about the traction control. That is a slightly different system from the stability control (VSC). Regardless, the reason it is so twitchy in the Prius is assumed to be protection of the electric motors.

Re:recipie for disaster

[tibit]

Care to explain how reducing engine torque (usually by cutting fuel supply) when the wheels lose traction causes accidents? Because that's what your cop buddy doesn't seemingly realize is all that traction control does. I think your buddy is just confused as hell, that's what I think, and you trust his confused self.

Re:recipie for disaster

[tibit]

People routinely confuse traction control with stability augmentation (a.k.a. ESC).

Re:recipie for disaster

[omnichad]

Based on what I've seen, they are. My experience was being stopped in 4-lane traffic, and an ambulance speeding by at 40 miles per hour BETWEEN the lanes!!

Re:recipie for disaster

[BlackSnake112]

With everything on, it makes you feel like you're the best driver in the world.

That explains one of the many reasons why the worst drivers I see are in luxury cars. They drive like nothing can stop them. The car will protect them in case of anything bad, so drive away and not care. I know not everyone does that. The greater number of luxury car drivers I see do drive like morons. That does not take into account I seeing more luxury cars also blowing more red lights, skipping stop signs, making a left turn from a right turn only lane, making a right turn from a left turn only lane, basically ignoring the big arrows painted on the road and going in any direction they want, riding the shoulder in traffic, riding the shoulder at a light and cutting back into traffic when the light changes that is if they even stopped or the red light at all. I see this happens every single day. Often by the same people. My favorite is they blowing a 4 way stop sign and yelling at everyone else that they are in the intersection and we have to stop for them. Maybe it is a north east cost thing. It has been the same in New York City, in Rhode Island, on Long Island, in New Jersey, in Philadelphia, in West Chester PA (south eastern PA areas really), In Delaware, In the DC metro area, northern Virginia, and Maryland from DC to Ocean City.

Re:recipie for disaster

[tibit]

Because, obviously, the wheels know they are subject to magical breaking torque from the engine, rather than ordinary braking torque from brakes, huh? If you did never lock up your drive wheels using engine braking, you haven't tried hard enough. Brakes are usually quite accurate and you can apply very minute braking pressures -- say decreasing your air-drag-and-rolling-friction-only (no feet on the pedals) braking distance by half. That means that it's quite feasible to apply only as much extra braking torque as is effectively provided by rolling friction and air drag alone. I wouldn't call it "light pressure on the brakes can be enough to lock the wheels". I can brake just fine on smooth ice without the ABS kicking in, if I want to.

Re:recipie for disaster

[Gordonjcp]

Right, but they do have rollover protection, which is not required in most US cars which is why they're so damn hard to insure over here.

Look at how American SUVs crush up like coke cans in even a minor prang if you want a good example.

Re:recipie for disaster

[tibit]

Yeah, but you see, in a constant pressure approximation like you claim, the tire is a balloon where surface is assumed not to carry any shear stress -- otherwise the pressure won't be constant across the contact patch area. Thus if you merely place such an imaginary tire on the snow, the pressure on the snow is determined by tire pressure. Nothing more, nothing less. But go ahead and test it on balloon-like-tires: smooth-ish wide bicycle tires are perfect for such a test. Rolling across your finger with 15psi in the tire is much more pleasant than with 30psi in the tire. On my wife's car, I merely squirm if she rolls over my foot with the front tire deflated to 15psi (I tested it before writing the answer, mind you).

Since car tires are not really shear-free balloons, the constant pressure approximation doesn't hold, and pressure is maximal in the middle of the patch, in absence of rolling. The ratio of maximum to average pressure depends on the circumferential length of the patch. Narrower tires produce a higher pressure peak in the middle since their contact patch is longer in circumference, having been forced to be narrower in lateral direction. Apparently that works better on the snow, but your "pressure = force x area" exclamation doesn't matter here, and grade school science usually doesn't talk much about pressure distributions in tires. You need to know a bit of mechanics of materials to understand what shear stress is and how a ballon-model-vs-tire differ in that respect. That's some very nice grade school you went to if they taught you that!

Re:recipie for disaster

[tibit]

Agreed somewhat. The assumption of contact patch area having a constant pressure is pretty much a grade-school kind of approximation. No street tire acts that way -- there's this pesky fact that pressure across the contact patch is not constant unless you're riding on a balloon. Real life tires carry significant shear stresses that allow them to produce a rather interesting non-constant pressure distribution across the patch. See my other post.

Re:recipie for disaster

[tibit]

That's not very accurate at all. You sort of get the idea, but are not sure at all how it really works.

Think of a balance of forces on a small segment of a tire. Air pressure is merely applied from the inside of the tire. If that's all there was, the tire would explode: after all, the forces have to be balanced and you don't have anything keeping the tire material from being pushed outwards by the air. If your tire is a balloon of sorts, very floppy, the air pressure will be counteracted by any external contact pressure, and by the hoop stresses in the tire material (the rubber!). So the stress in the rubber most definitely has to counteract all of the air pressure, except when there's external load on the tire! This holds true whether the tire is floppy or not, only that stresses other than hoop stresses and contact stresses come into play.

What you mean by "the rubber carries the remaining weight" is, in reality, the fact that the tire not only carries the normal stresses, but also shear stresses. And it's not only the rubber doing that, there are polyester and steel wires (usually) helping out! When you add shear stresses into the mix, then you can deviate from the constant contact patch pressure. A modern good car tire has, IIRC, about 40 rubber components of various properties, so it's not just a monolithic cast rubber chunk with some wires inside!

Your claim that "setting n to be large" is pretty much meaningless, there's no such "n" in any sort of tire engineering. You could use it for some sort of a pretend "advanced" tire model that's only fit to use in grade school. If you want to go that far, I'd much rather introduce the conept of free body diagram (if not already done) and shear and normal stresses, and use that to realistically show what happens in a tire. Your point of view is pretentious in that it seems to be quantitative, but the quantity doesn't describe any behavior at all.

You could, perhaps, argue that "n" is the ratio of shear-to-normal stress, but then you have to be explicit as to where in the tire it's checked, under what conditions, and what shear and normal stresses are talked about (there are several you can choose from!) -- so again, not a useful figure the way you describe it, and not a figure that's used in any sort of engineering!

Re:recipie for disaster

[tibit]

Electric lights are dangerous and a bad idea. Gas lights are dangerous and a bad idea. You get the drift of that, I'm sure. Informative, my ass.

Re:recipie for disaster

[RobertLTux]

the trick is that if you start with a wider gene pool to start out with then you have a greater chance of having the "dice" find something interesting.

look at it this way if you have to roll 35 on 4d10 to evolve and you start out with 8d10 its a lot easier to do this.

(btw this shows a logical fault in evolution since way to many things have A-MG chains of traits that if they don't all work then

GAME OVER (You Have Died))

Re:recipie for disaster

[cayenne8]

By that time, computers will likely be so sophisticated that it would be considered a crime against humanity to drive a 2 ton vehicle manually.

I hope that day comes LONG after I'm dead and buried....

Driving a high powered sports car is one of life's fine pleasures!!!

Re:recipie for disaster

[tibit]

modern vehicles have two independent hydraulic systems going to the calipers

LOL, you're making shit up. Even very safety concious car makers like Volvo don't do that, and IIRC never did! Older Volvos used to have independent front and rear brake circuits IIRC, and there was an isolation valve so that should one of them produce more flow than the other, the high flow one would be cut off and the other one would still be operational. When this valve tripped, the BRAKE idiot light would illuminate. I don't recall two brake lines going to calipers even there!

On a modern car, the brake circuits are individual between the brake control module (a.k.a. ABS module) and the calipers, but there is just a single brake line going to each caliper! There is no mecahnism implemented that would detect a leak in an individual brake line and plug it. There are actuators to do just that, but there are no flow sensors, and in case of typical pinhole leaks, the brake pressure is only lost when you are out of brake fluid, so merely sensing circuit pressure wouldn't help you.

I have had a pinhole leak in a brake line in a 2000 Volvo, and sure as heck it'd leave you without hydraulic brakes. A friend of mine had the same on a nice 2004 BMW, with same outcome. You're really talking out of your ass, and you well know it. Someone please moderate the parent as a troll, because he sure as heck deserves it!

Re:recipie for disaster

[ultranova]

The stupid people breed a lot more, so the average intelligence of the population would plumet.

Aparently.

Re:recipie for disaster

[Quietust]

If you did never lock up your drive wheels using engine braking, you haven't tried hard enough.

Last I checked, "wheel lock up" means the wheels cease rotation and start skidding uncontrollably, so the only way you could possibly lock up your drive wheels with engine braking would be if you stopped the engine - as long as it's still running (and the transmission is engaged), the wheels will keep turning (though they won't provide much torque unless you're driving an automatic and you're at a complete stop).

I will agree, though, that strong negative torque from engine braking (equivalent to what would cause your brakes to lock up the wheels) can definitely cause you to lose traction and start skidding, but it won't lock the drive wheels unless you define locking differently.

Re:recipie for disaster

[eth1]

No, electronic stability control is a great idea that makes cars much safer ...for most drivers.

For highly skilled drivers, it probably gets in the way more than it helps. But the vast majority of drivers aren't pros.

Even if you are a pro, the street is not a track. I have a car that I track occasionally, and the ESC *does* get in the way there, if I don't turn it off. But that only matters when you're close enough to your handling limits that you can use that wheel slip to your advantage, AND know ahead of time that you're going to do so. On the street, driving that close to your limits would take you from "pro" to "dangerous idiot," and when the time comes, you're not going to have time to study a map of the corner/avoidance manuever to plan the best line through it, or any practice laps.

ESC goes off at the track, but never on the street.

Re:recipie for disaster

[ultranova]

I've been in rollovers in vehicles with roll cages,

The plural here makes a rather strong case that your car should be driven by the computer to as great extent as possible, for the sake of other road-users.

Re:recipie for disaster

[LateArthurDent]

This, and increase your braking distance. Riding up the guy in front's chuff in ice is asking for a big hike in your insurance premiums.

Where I live, snow can sometimes happen, but we don't get it every year. As a result, people have no idea how to drive in it. Well, 12 years ago I was in college and it started snowing right as I left class (so they didn't cancel it, like they usually do at any sign of white stuff). So I got to my car to head home and everybody was trying to get home as quickly as possible. I'm not going to say I'm a great driver or anything, but I knew enough to keep an increased distance between me and the car in front while on the interstate. Well, impatient drivers don't like that, and traffic is already moving at 20 mph. So this guy in a truck with his friends honk their horn at me while trying to thread his way to the lane next to me, gets on the other lane, flicks me off while passing me (for probably a full minute, it's not like his lane was moving that much faster), cuts back in front of me and proceeds to immediately hit the brakes and slide into the rear of the car in front.

If it weren't for the inconvenience of the poor guy who got hit as well as everybody else who now has to deal with the increased traffic caused by the accident, I'd call it justice. I have to admit to laughing when it happened, especially as I passed him. Lesson of the day: don't be an impatient douchebag, you'll get home quicker if you drive smart.

Re:recipie for disaster

[dotancohen]

I never said that there are two lines going to each caliper. On line is split between front right and rear left brakes, and the other is split between front left and rear right brakes (be the rears discs or drums).

I'm actually interested in hearing about the pinhole leaks that you and a friend had on two different vehicles that left you both without hydraulic brakes. Such a leak is not unheard of, but I've never heard of it leaving a vehicle without a hydraulic braking system. In the normal scenario for that type of failure, one hydraulic system is compromised. In the worst case scenario brake fluid would leak out and air/moisture would leak in, leading to inconsistent performance which may trigger ABS / ETC systems.

Re:recipie for disaster

[dotancohen]

Thank you, I was summarizing but I appreciate your in-depth review! I am actually not familiar with the engineering aspects of shear stresses on the tire while cornering, only with straight-line acceleration and braking, so I was not referring to that.

Re:recipie for disaster

[dotancohen]

I would very much like to hear more about this. Are your sure that both hydraulic systems were compromised?

Re:recipie for disaster

[ircmaxell]

Yes

Basically, it used one valvebody with two separate chambers (one for each system). However, the main valve cover covered both chambers. So when the cover's seal blew, both were compromised. Granted, it's an edge case. But it did happen to me...

Re:recipie for disaster

[rjr162]

"Additionally the best seat belt in those cases is a proper 4-point harness"

Make that read 5-point instead of 4-point and I'm on board.

But to add to your comment, if you want to solve the neck issues, I guess every car should also comes with a HANS device

Re:recipie for disaster

[danomac]

It's possible to skid with engine braking (as I found out the hard way!)

If you have a manual transmission and choose the wrong gear (I accidentally went to first instead of third) the rotating mass may be too much for the engine to handle and it will lock everything: engine, transmission, drivetrain, tires and you go into a skid. :(

Re:recipie for disaster

[rjr162]

Of if your area doesn't allow for studded tires, get some Blizzaks as mentioned above. They are made with a special process that introduces tiny bubbles into the outer layer of rubber, which as it wears makes a "suction cup" type area to help grip the road (so there's like millions of suction cups contacting the road).

Another note to mention, winter tires in general have softer rubber compounds, which allow them to flex more and grip better in the colder weather (also the sipping helps a lot as well). Now the rubber compounds in your longer life all-seasons and what not are a harder compound that won't function as well in the colder weather.

Same principle used when people Solo II race. The Falken Azenis tires are cheap for what they are and are a softer compound. This means a shorter life than a normal all-season, but increased grip on the track (especially for Solo II where the courses aren't large and you get 3 to 5 trips around, so the tires don't have mad warm-up periods). The downside is if it's too warm out or you run them too hard, they can become "greasy" as the rubber compound is now too warm and starts slipping again.

Re:recipie for disaster

[danomac]

My point was that when you introduce electronics into the mix it's hard to predict failure modes. Electronics can fail in pretty darn bizarre ways, as I'm sure you've seen.

While augmenting a lot of safety systems with electronics is the norm, it's far harder to do failure mode testing. If voltages or sensory inputs are a little bit off you may be in for a bit of a surprise.

In the case of the car I was talking about, it was a faulty sensor. It apparently failed in such a manner that the engineers had no fail safe for that particular type of failure. So basically, the electronics prevented me from applying the hydraulic brakes, which were in perfect condition with no issues. Had there been a car or a person in front of the car when it failed, they would definitely have hit, and in the case of a pedestrian, been in the hospital.

To those that say use the emergency brakes, it's not practical, especially if you're trying not to hit something. That car had a ratcheting brake system and required three or four pumps of the pedal to apply. That's how I did eventually stop, but it would not have helped me. It took ten seconds or more to apply it, and then it very slowly dragged off to a stop (about 30 metres.) I rolled right through an intersection before coming to a rest. This was at 10kph.

So yeah, the parking brake can stop you if you're in no danger of hitting something. How often does that happen? There's almost always something in front of you.

I'm not saying safety features are bad. Sure, it helps protect 99% of the people that don't know how to drive. It really sucks if you're in that 1% and it does something you have no control over.

Re:recipie for disaster

[Carewolf]

Besides, a good abs system won't trigger unless you lock the brakes

Or you are driving on snow or gravel. The computer can not tell the difference between tires skiding due to near aquaplaning, or skidding due to driving on a loose surface. Any ABS system will go out its way to try to kill you if you do not disable it before driving on snow. Note some ABS system has a 'snow' mode, which is just a safer and more friendly sounding name for off mode.

Re:recipie for disaster

[T-Bone-T]

I've seen the same thing happen to a pair of rich kids with a BMW and Mercedes on an elevated and curved off-ramp.

Re:recipie for disaster

[fnj]

Back to driving school for you.

Re:recipie for disaster

[Ironhandx]

I didn't miss the panic-braking. It is specifically discussed. When someone panic-brakes they are panic'd . They nearly invariably death-grip the steering wheel rendering all mobility gained from the ABS moot.

My friend is someone that used the mobility, but he learned to drive on ABS, he is the type of person that could easily have been braking properly himself.

Re:recipie for disaster

[Ironhandx]

That Volvo has better safety belt systems than 90% of the vehicle systems sold in North America in that case. There is only a mechanical locking mechanism on the majority of the cars sold here.

Re:recipie for disaster

[Fernando Jones]

this sounds eerily reminiscent of people who like to smoke in bars....

Re:recipie for disaster

[Ironhandx]

What you suffered is the 90% case for air bags going off. You have to be in awkward or otherwise in mid motion with the air bag going off in order to generate most of the other injuries.

What causes the problem with seat belts is it will throw your head forward at additional speed, your head will then rebound off the air bag that much harder. Its generally the impact with the additional acceleration causing tears in the neck or the rebound doing the same that causes the injuries.

Re:recipie for disaster

[dotancohen]

I'm not familiar with the component, but some master cylinders had a single seal between both systems. When the seal blows, the only side effect is an equalisation of pressure between both systems. This is not a problem unless one of the systems also has an independent leak. Might this have been the case on your Chevy as well? Were any other repairs made, other than the valve body?

Thank you for letting me know!

Re:recipie for disaster

[Ironhandx]

Off roading. I've never even been in an accident on-road.

Re:recipie for disaster

[Ironhandx]

Those are some excellent stats. Thank you sir.

They certainly can be, but I still think it should be a personal choice. My statement isn't that seat belts are universally bad, its that they shouldn't necessarily be mandated, because there are many driving situations in which it is not a good idea to be wearing one. I probably experience more of those than most because of where I live, but the fact remains the situations exist.

I'm fine with mandated rules for baby car seats etc because they at least advise a best practice as the baby is not able to make its own decision. I'm fine with legislation up to age 18. After that it should be a personal choice.

Re:recipie for disaster

[cayenne8]

this sounds eerily reminiscent of people who like to smoke in bars....

What's wrong with that....?

And aside from some states...you can still smoke in bars as far as I know...at least we still have that freedom of choice here where I live....

Re:recipie for disaster

[cbeaudry]

The ATTESA is an AWD system, not a traction control system.

Now AWD systems do incorporate traction control, as they need to electronically or mechanically send power forward, or to the back and account for slippage.

Re:recipie for disaster

[cbeaudry]

I drove my RX8 in Montreal winters for 5 years. What you're saying is just untrue.

I found it to have one of the better traction control systems I have driven in. The corrections where minor and always done at the right time.

When you hit a patch of black ice, there really isn't much any system can do. If you went into a spin, its because you made a steering correction, or you weren't going straight when you hit the patch of ice.

I would constantly test the RX8's system in snow or ice, Id give it a sideways swing, step on the gas and keep the steering straight ahead... the car would swing "lightly" back and forth (because of the aggressive input steering I had given it), but its trajectory would remain STRAIGHT as long as I kept the steering straight.

After a few swings it would eventually taper off and just drive straight.

Honestly, DSC/TSC systems are a burden when on a race track, but they are not on the road, even for experienced drivers. (I do Solo I and Solo II, time attack and other road racing/lapping activities) and I still keep DSC-TSC on when I drive on public roads.

Re:recipie for disaster

[ThatsMyNick]

Well, we have evolved to hold humans a special status. Even new born children can differentiate humans from other creatures/objects, and treat them differently. Though, some people are just different I guess.

PS: I wouldnt have modded you troll, but I would certainly have categorized you as strange.

Re:recipie for disaster

[green1]

I have had this exact problem on all 3 of my north american vehicles. right from brand new. I doubt it's a faulty sensor. I think it's more poor design.

Re:recipie for disaster

[green1]

According to the decal on the windshield this truck was proudly made in Ford's Kentucky plant. I also had the same issue on my last Ford, and the Chev before that. My Mitsubishi, while still worse than no ABS, is much much better and has never come close to causing a collision.

Re:recipie for disaster

[green1]

Good, if we have the choice I'll take me driving, and you'll take computer driving, and I'll continue to drive regardless of the conditions safely, and you... well, you take your chances.
What I can't tolerate is the idea that they'll take the choice away from me in the name of safety when we have proven several times that with even a half competent driver most of these safety features work worse than when you don't have them.
You want to fix safety on vehicles. Start with who you let drive. As long as we refuse to make drivers pass proper testing before getting behind the wheel, and as long as we refuse to stop people who have proven repeatedly that they can't drive from doing so, safety will never improve, regardless of how many gimmicks you put on the vehicle.

As for "caring about perceptions" I care about reality, not perceptions. And my ABS has nearly caused several collisions in no imaginary way. I drive vehicles both with, and without ABS, I never have these problems without ABS, but with it...

Re:recipie for disaster

[green1]

Bingo!
Mitsubishi ABS... bad, but about as good as ABS gets
Ford ABS (and the Chev one I had before)... downright DANGEROUS. I can't believe they are allowed to sell such a thing.

And that doesn't even go in to the fact that there are times where locking up the wheels is the right thing to do to avoid a collision. (of course proper threshold braking is also better than ABS every time)

Re:recipie for disaster

[green1]

Except that if only 1 wheel is sliding, I want all 4 wheels to continue to brake normally. because only 1 wheel sliding means that the surface that's slippery is small enough that it doesn't affect much. In addition to all 4 (or at least both front which is where most of the braking power is) wheels doing the ABS thing and loosing most of my braking power, they also do it for far longer than necessary, I pass over a manhole cover, and the ABS is going for the next 30-50ft even though the manhole cover was only 2ft accross. meaning I'm standing on the brake and passing 50% of the way through the intersection even though without the ABS I'd stop before the stop line.

Re:recipie for disaster

[holophrastic]

No, that's not enough. Just because they save the lives of 10'000 bad drivers, I don't want it to kill me if I'm a good driver. It's not about results. It's about accountability.

Re:recipie for disaster

[adolf]

If I had a vehicle acting in this fashion, right from brand new, I'd have had it fixed. And if it couldn't be fixed, I'd have it lemon lawed and buy something else that actually worked.

Because frankly I only see two possibilities here: You've been unluckily stricken with a series of dangerous new vehicles, or your driving practices are garbage. And I'm trying to give you the benefit of doubt.

Re:recipie for disaster

[green1]

Maybe you missed how my personal vehicle behaves much better. So it's unlikely my driving style.

I think most people put up with vehicles that in a bad situation make things worse instead of better.

Re:recipie for disaster

[adolf]

In that instance, neither of us are "most people." We needn't include them in this discussion.

Whatever the case, I'm intrigued. Feed me some model/year data on the cars in question and I'll try to dig up who actually built the ABS systems.

(Please note that car manufacturers normally don't handle this on their own. For instance, my praised E36 BMW that I helped start this thread with, waaay up there, has a braking system (including ABS) designed and built by Ate/Teves -- not BMW. Generalizing ABS behavior by the badge on the front grill is a recipe for vagueness and unknowledge and leads to confusion instead of a well-grounded understanding.)

Re:recipie for disaster

[drinkypoo]

Back to driving school for you.

Back to automotive technical school for you.

Re:recipie for disaster

[adolf]

And, but, so? BMW doesn't make FWD cars under their own name, and I'm not familiar enough with Mercedes to know (except they seem to be primarily RWD), but: What, exactly, is your point other than to be uselessly snarky?

FWIW, the car in question was a 1995 Chevy Beretta in its most barren trim/accessory package. $13,995 new off the dealer lot, or somesuch, and certainly not any manner of high-end kraut-wagen: Friggin' 14" wheels with plastic covers and rear drum brakes.

Weight distribution on a Beretta is an ugly joke. Accordingly, one wouldn't expect gravity to be able to pull the rear end downhill in any sensible circumstance, but it did.

(As an aside: I currently drive a E36 BMW, but I'm certainly not rich -- I don't even qualify for "middle class" by a long shot. If you think only rich kids can drive a BMW, you haven't seriously looked at owning one yourself.)

(As another aside: Perhaps I should just killfile you as a troll. Why not?)

Re:recipie for disaster

[T-Bone-T]

I misread fwd as rwd. Anyways, my point was that these kids were killing their only source of power by using too much while I drove right around them in my fwd car.

Re:recipie for disaster

[jamesh]

The world is full of engineering disasters that occured because the designer missed one little thing.
Such mistakes improve the world of engineering by learning from mistakes.
But peer review also helps.

Peer review is great, but shouldn't be conducted on what is essentually a press release.

Re:recipie for disaster

[jamesh]

No citations (except those on Google's approach to such problems, which may or may not be relevant here so I won't provide them. I'm sure you can find them on Bing :). I've got no idea how the designers resolve this, but I don't really need to to reach my conclusion - for the PP's proposed sequence of events (running another car off the road because a tumbleweed blew by) to take place on a system certified for use on public roads would require a catastrophic series of failures, including, but not limited to:

. None of the hundreds of people involved in such a project being aware of the possibility of objects existing on the road that pose no risk in a collision (tumble weeds, plastic bags, rain, etc)
. Nobody encountering such objects while conducting testing
. Nissan's legal team not imagining such a failure
. None of the certifying bodies doing any sorts of testing

And that's only at a superficial level. And even if all of the above and more went wrong and the broken feature made it into production cars, such a failure would be encountered within hours of the first such car being driven off the lot, at which point the whole range would be recalled. I just don't see how it's possible.

Re:recipie for disaster

[ultranova]

I will agree, though, that strong negative torque from engine braking (equivalent to what would cause your brakes to lock up the wheels) can definitely cause you to lose traction and start skidding, but it won't lock the drive wheels unless you define locking differently.

But since wheels locking is bad precisely because it causes you to lose traction and start skidding, the advantage of them doing so without locking would seems questionable at best.

Re:recipie for disaster

[samwichse]

That is why for many years (before ABS), manufacturers would put disk brakes on the front but old-style drum brakes on the back. Because IF you are going to get a lockup, you want it to be your front wheels, not the rear.

Except that this is not true. Drum brakes are much more prone to lockup than disk brakes. Drums don't just brake using the hydraulic cylinder that presses the shoe into the drum surface. The drum is rotating, and the twisting motion of the shoe mechanism gives it additional force without having to press the pedal (why cars could get away without brake boosters when they were all 4 wheel drums). This effect is called "self applying." It's why cars with rear drums have more effective parking brakes (rear disk parking brakes SUCK).

http://en.wikipedia.org/wiki/Drum_brake#Self-applying_characteristic

Sam

Re:recipie for disaster

[Vlado]

The thing is that ambulance drivers are above-average trained drivers.

I would say that majority of people on the road have absolutely no clue what to do if the situation occurs that their car loses grip on the surface. Most of the time they will simply either slam on the brakes or do absolutely nothing. And most of the time the end result will be crash, with consequences proportional to the initial speed with which they were going.

ABS is usually considered a good thing in the car. But a lot of the drivers will tell you that if you have a feel for when your car looses grip, when you break, you'll break more efficiently without ABS than with it. In order to do that, however, you have to test emergency breaking scenarios a few times. If not: Serious shit -> full slam on the brakes with no ABS -> probably crash after you watch the obstacle in front of you coming closer and you're without any control over you'r vehicle, that you're aware of...

For a vast majority of drivers "crap" like ABS, ESC and the like will be life-savers. For some above-average, they will be a hindrance or a danger.

Re:recipie for disaster

[green1]

Personal vehicle (acceptable ABS):
- 1995 Mitsubishi Delica L400 Spacegear

Work vehicles (dangerous ABS):
- 2001 Chevrolet Astro Van
- 2006 Ford E250
- 2008 Ford F350

Re:recipie for disaster

[tibit]

Shear stresses in the tire material have got nothing to do with where you are driving, and cornering is absolutely irrelevant to this discussion. Don't mix up shear contact forces with shear stresses. Those two got nothing to do with each other!!! Remember: a stress is a property defined at a point in the material, and there's generally 6 numbers needed to completely describe stress state at a point (plus another 3 numbers to describe what orientation is your stress state described in).

Re:recipie for disaster

[tibit]

OK, I missed that important detail. There's no magic in keeping the wheels rolling while slipping, though. Once there's slip, it doesn't matter if they're locked up or not. Dynamic friction doesn't care much about what's the velocity difference -- once there's enough of it, it doesn't matter anymore. Of course keeping the wheel rolling means that recovery from the slip will be faster than if the wheel was locked up: if the friction force is constant, then it'll simply take less time to transfer the smaller energy needed to speed up a wheel that has non-zero rotational velocity. Decent ABS systems try to maintain such limited slip, thus the more slippery the surface you're on, the higher pitched will the brake modulation be, as brake force gets to be applied for shorter periods of time. Best ABS systems use periodic modulation only for measurement of slip and maintain arbitrary braking force, adjusted by the results of the said measurements. It's like if you applied just enough brake pressure, and wiggled your foot a bit to see if you could apply more by any chance. I don't know offhand who does it that way, as it'd be considered state of the art at the moment.

Re:recipie for disaster

[tibit]

Once there's slip, it doesn't matter if they're locked up or not.

Well, some special muddy/snowy conditions notwithstanding. That's why it's silly that there aren't ABS mode switches that'd tailor the response to actual conditions. I guess with some idiot drivers unable to turn their lights on correctly, that'd make things worse, not better. What I mean is that all too often I see people driving at night with only position lights turned on :(

Re:recipie for disaster

[ergean]

9 and 3
8 and 4

But like a stupid person I keep my hands at 10 and 12 - I have no idea why I do this. 12 or the parking brake.

Re:recipie for disaster

[adolf]

So. Your point is that kids are still just kids, even in these modern enlightened times.

Got it.

Hmmm...

[Nexion]

I bet it will be just AWESOME on ice!

Re:Hmmm...

[Anonymous Coward]

I wonder who the first fucker will be to hack together a radar/laser jamming system that sends these cars onto random uncontrollable trajectories and sets it up by the 405 at 7:30 in the evening?

Re:Hmmm...

[nospam007]

"I wonder who the first fucker will be to hack together a radar/laser jamming system.."

It's called a 'ball'.

I hope

[Cute and Cuddly]

I hope that there will be plenty of logs, just in case that when your car avoids a dog and kills a kid you can go to cort and blame Nissan for it

Re:I hope

[Jane Q. Public]

"I hope that there will be plenty of logs, just in case that when your car avoids a dog and kills a kid you can go to cort and blame Nissan for it"

You nailed it. Because that's the typical history of this kind of technology:

1. It is introduced.

2. Somebody sues on the grounds that it caused an accident, rather than avoiding one.

3. It is taken off the market.

4. It is gradually improved, and finds its way back into the market. Fairly typical time frame: 10 years.

Re:I hope

[dunng808]

Examples ...

Re:I hope

[Jane Q. Public]

Air bags, antilock (ABS) brakes (note: that is a report from back in '98), steering wheel locks (too old to find easily on Google)... etc. Just about everything that has had to do with removal of operational control from the driver.

Re:I hope

[Cute and Cuddly]

If they use GNU/linux, there will be plenty of logs by default in /var/log. Adding logs for procedures is kiddies play

Re:I hope

[Jane Q. Public]

Almost forgot... all the famous cases having to do with computer-controlled cars. Many, many lawsuits. Relatively slow adoption. Mistrust leading up to the present day.

Re:I hope

[SleazyRidr]

Toyota Prius acceleration issues?

Re:I hope

[samwichse]

Were proven to be BS.

Bugs

[multiben]

Programmer accidentally casts unsigned to signed: steering wheel turns full lock right instead of a bit to the left?

Re:Bugs

[multiben]

Yeah, I was sort of having a bit of a joke. But thanks for lecture!

I, for one,....

[jimbouse]

I , for one, welcome our new automatic steering overlorrrrdd....... Wait, no I don't!

Traffic accidents happen fast. Normally due to 2 distracted (or impared) drivers crossing path.

If you are a defensive driver, you always have an "out". I, like a responsible adult, keep my distance, travel at a safe speed, respect the road conditions, etc. I have only had one near-miss and it was due to someone running a red-light.

Re:I, for one,....

[Velex]

I disagree. In order for this to work, you have to have an out. I'm no luddite when it comes to automated cars. Automated cars will save lives. But this system makes me wonder.

An automated car will be leaving itself an out, making sure other drivers see it, and using all the other keys of the Smith System with more precision than a human could hope for. Is this system tracking all those things too? Is this system up to par with Google's?

Or is this system going to send me into another car because a squirrel runs in front of me and it's only tracking what's directly ahead of me?

I am wary of these

[Gaygirlie]

I cannot help but actually be quite wary of how these kinds of things are starting to crop up in modern cars. How well do these work e.g. in an abnormal situation, like e.g. there's flood water on the road, or lots and lots of snow like we here in Finland tend to get, or what if the system detects something on the road that it really wants to steer away from, but doesn't detect what's coming up on you from the side and steers you in even worse direction against your wishes? Hell, a proper snow storm is a common occurrence here in Finland and even humans have trouble keeping track of everything that's happening; I really, really doubt a computer can do a better job at it.

These things might be good for people with serious attention deficits or other kinds of similar issues, but an experienced, careful driver could quite possibly make better decisions than these and thus these systems would actually be detrimental in such cases. Hell, they could just as well turn a not-so-serious crash into a major disaster if they screwed up and took control of the car over the driver.

Re:I am wary of these

[Anonymous Coward]

These things might be good for people with serious attention deficits or other kinds of similar issues, but an experienced, careful driver could quite possibly make better decisions than these and thus these systems would actually be detrimental in such cases.

The key phrase in what you said is "an experienced, careful driver." Unfortunately the majority of drivers on the road are not experienced, careful drivers. Too many of them are driving beyond their abilities or just doing plan stupid shit - like the police officer I monitored on the freeway this morning, trying to drive the cruiser while he used a mobile phone in one hand and managed a cigarette in the other. Why do you suppose he was having problems staying in his lane? The proposed system might actually help. Maybe.

Re:I am wary of these

[HideyoshiJP]

I agree that there's a huge lack of decent responsible drivers on the road. Unfortunately, I can't help but feel like this is the wrong approach. Instead of fixing the drivers and punishing the ones who don't improve, this simply enables those who can't drive well to stay on the road. I will admit that I'm heavily biased, and an idealist, so my opinions on this matter should be taken with a grain of salt.

Re:I am wary of these

[notdotcom.com]

Have you ever considered that these systems might be connected to other systems?

If you're in a low area of Florida, near a coast, river, ocean, etc (determined by GPS), during a wet/hurricane season (determined by date/time, humidity and temp sensors), crossing standing water (determined by moisture, humidity sensors, on-board cameras, input from differential and traction control sensors, engine speed, load, and RPM, emergency warning systems via radio), and going less than 25 MPH (speed sensors, GPS, cameras, etc)... then perhaps the system could be bypassed for those occasions when you needed to cross a foot of water over the road.

Maybe it also just has an "off road" mode that would disable it (after a nice lawyer screen).

Re:I am wary of these

[robot256]

Given that there are only a finite number of cops on the road, and some offenses are difficult to detect let alone stop someone for, I assume that means you are in favor of more automated enforcement? Like red light cameras, don't-block-the-fucking-intersection cameras, insurance company dashboard boxes, etc. There are only two options to improve safety: make driver's licenses more like pilot licences in rigor, or make the cars drive themselves. Nissan just got closet to the latter. Just like Google said, self-driving car technology will be introduced slowly until eventually the cars *can* drive themselves.

Re:I am wary of these

[HideyoshiJP]

You're absolutely right, and I definitely lean toward a hardened licensing process. I also support higher fines/penalties. Unfortunately, these will not realistically happen.

Re:I am wary of these

[AmiMoJo]

You could make the same argument against driverless cars, but they seem to be pretty safe. Nissan are using that same technology, complete with radar and cameras to make sure it only engages if it can be sure that the road is clear.

We are now reaching the point where computers can drive better than we can.

Re:I am wary of these

[Solandri]

Here's my abnormal driving situation story. I had just taken the off-ramp from the freeway when the radio started acting funny. Being a gadget geek, I set about trying to figure out what was the problem. A few seconds later I suddenly remembered, "Hey, I'm supposed to be driving!" I looked up to see myself about to plow into a car at the bottom of the ramp stopped at a red light.

I slammed on the brakes, but due to the ramp going downhill it wasn't going to stop me quickly enough. To the right was a guardrail and a ledge. To the left was a curb, a small dirt lane, then the wall leading up to the freeway. I made a split-second decision that the dirt lane was my best choice. I spun the wheel left, hit the curb, blew out two tires and destroyed a third, jumped the curb, then jogged the wheel right. I made it onto the dirt, avoided the wall, and stopped beside the car I had been about to hit.

Thinking about how an automated system might handle the same scenario, I can think of multiple ways it could fail. First the ramp was downhill, and so my braking distance was increased. An automated system might get confused by that because a simple accelerometer cannot distinguish between angle of the ramp and deceleration. You need a gyroscope to correctly determine orientation. Consequently, even if the car had automatically applied the brakes when it sensed me approaching the stopped car, it might have already been too late.

Second, the dirt path wasn't big enough for my car - I ended up two wheels up, two wheels still on the off-ramp (that's actually probably what kept me from hitting the freeway wall). Fortunately the car stopped at the light was in the middle of his lane, not glued to the left. An automated system would probably try to (1) prevent me from driving up a curb, (2) refuse to aim the car towards a wall, and (3) prohibit entry into a dirt lane too small for the car's width.

Finally, given the obstacles to my front and left, an automated system may have been confused by or decided that going off the ledge to the right was the best option.

This isn't to say that these sorts of problems can't be overcome by a computer. It's just to point out the convoluted sorts of trade-offs a computer may be forced to evaluate when faced with an "abnormal" driving situation. I do not envy the programmer tasked with developing generic accident-avoidance routines designed to cope with all abnormal situations.

Re:I am wary of these

[digitig]

It would have to be a lot better than the average driver. A few reasons:

• People are far far more willing to accept risks that they control than risks they do not control, and are more willing to accept risks due to human actions than due to automated systems. That is one reason they are happier driving than flying, even though flying is much safer. I recall being at a safety conference at which there was discussion of a robotic system for doing prostate operations. The most serious failure mode was leaving the patient incontinent, and the research showed that the robotic system would have to be more than ten times more reliable than the human surgeon before it would be accepted by the public.
• It shifts legal liability. Even if the system prevents a million serious accidents for every one it causes (which would be a great benefit, of course), the legal liability for that one accident would land squarely on the manufacturer, whereas the accidents it would have prevented would come down to the drivers. The manufacturers are going to be reluctant to accept that liability.
• As soon as there is a serious accident caused by a system error then (no matter how many serious accidents the system has prevented) there would be media outrage and massive lobbying to have the system outlawed.

False positives

[AaronW]

I can just see it, a metallic balloon drifts through a lane of traffic and the Nissan goes into panic mode and starts a big chain reaction because the radar, camera and laser scanner detect it as a threat. A real driver would just try and pop the balloon.

That is also one of my questions about how the Google self-driving cars behave in similar situations. Do they panic when a tumbleweed blows across the road?

Re:False positives

[profplump]

First, it's dangerous to drive into any obstacle. You do you know that box/balloon/etc. doesn't have a bowling ball in it?

Second, why do you assume the system would be any worse at choosing among the options -- left/right/straight -- than a human would? There aren't a lot of details on this system, but there's no reason to believe it doesn't *also* check beside the car before telling you which way to turn.

Re:False positives

[RandomFactor]

"a metallic balloon drifts through a lane of traffic and the Nissan goes into panic mode and starts a big chain reaction"

Red ones are the worst about this... http://www.youtube.com/watch?v=14IRDDnEPR4

Re:False positives

[Anonymous Coward]

What atmosphere do you live in that a regular balloon can lift a bowling ball? That's some dense air!

Re:False positives

[TheLink]

I think the real density is elsewhere ;).

Re:False positives

[TheLink]

I don't think they intend to put this on the market yet. It's just one of the steps they need to take for a more automated car.

Re:False positives

[scdeimos]

Are you perhaps assuming that the fixes for their concerns weren't vetoed by their managers or nixxed by the cost accountants upstairs? Coz, like, that never happens in the real world.

Re:False positives

[JD-1027]

These types are arguments are great for discussion, but we need to always keep in mind that it only needs to be safer in aggregate not for every case imaginable.

NAITO RAIDAAHH!!

[sanman2]

They were all glued to the boob tube back in the 80s, soaking up inspiration for future products:

http://www.youtube.com/watch?v=ytJDJ96Es8Y

Predictable...

[kiwimate]

If Slashdot had been around 120 or 130 years ago, the first story about a new invention with four wheels and a motor would've been rife with comments about pedestrians not being able to get out of the way, horses being frightened, and predictions of other problems so severe the automobile would never see the light of day as an invention for the common man.

Re:Predictable...

[DNS-and-BIND]

Those were all real problems with the first automobiles. It was about 30 years from Karl Benz to the Model T. We are still firmly in the "Stanley Steamer" era of automated cars. Pedestrians and horses were well-advised to avoid such monstrosities.

Re:Predictable...

[Burpmaster]

No legs. Less clop than a horse. Lame.

Re:Predictable...

[Wordplay]

If Slashdot had been around 120 or 130 years ago, they'd have *invented* the thing with four wheels and a motor so that they'd have something with which to make analogies.

What could go wrong?

[dunkindave]

I mean really, I have complete faith in those automotive engineers to have envisioned every single possible condition the system will have to deal with. And also to have designed the electronics so that even if a component or wire fails, hell, even if a bunch of them fail, the system will automatically do the right thing. Don't you?

Re:What could go wrong?

[Anonymous Coward]

I have complete faith in these humans to envision every possible condition they will have to deal with, and to have the reaction times and attention to the road needed so that even if any component of their car fails, they will always automatically do the right thing. Don't you?

Robots!

[preaction]

You are experiencing a car accident! Do you require assistance?

Re:Robots!

[Tastecicles]

FUCK OFF CLIPPY!

Re:Robots!

[Migraineman]

"Citizen, you are about to experience an accident." [auto-drive initiates "cause accident" subroutines]

How much force does it turn with?

[dubbreak]

I was recently in a car accident. An elderly man drove into us from the front passenger corner. We had nowhere to go as there was a tall meridian on the driver's side. Basically a case of: scrub off as much speed as possible by emergency braking and brace for impact.

When we were struck the steering wheel was forced to one side. Much like the auto system in the video. The difference was I didn't know that was going to happen (like the testers) so I didn't let go. The spinning wheel resulted in a broken radius and I now have a large plate and many screws in my arm holding it together. Are you expected to let go of the wheel when the alarm sounds? My guess is most people won't be able to react like that. No problem for the tester since he knows it's going to occur, in real life you most likely won't be able to react that quickly or your reaction will not be to let go.

Personally I'm a fan of systems like Mercedes has that primes the brakes for you when a possible collision is detected (so it's easier to apply full brake force). Auto braking is OK too, but there are too many factors in auto steering imho. Lane correction is one thing, this type of emergency steering is a whole other beast.

Re:How much force does it turn with?

[reboot246]

Tall meridian? Was he from Meridia? I thought people from Meridia were very short.

Where I live we have medians, though some people insist on calling them "mediums".

Re:How much force does it turn with?

[dubbreak]

Yes I meant median

Re:How much force does it turn with?

[dubbreak]

Yeah, drive by wire would have saved my arm. That being said I'd prefer a hybrid system. It's nice to have proper road feel (go over a bump and you feel it in the steering wheel), but if some outside force moves the wheel dramatically then it decouples (sensing where the torque input is). Implementing such a system in an electric assisted steering assembly actually wouldn't be that difficult (often the wheel inputs to the electric assist which outputs to the rack, so it's just a matter of quickly decoupling the steering from the electric assist).

Deadful idea...

[Anonymous Coward]

An extremely awful idea.

The best collision is head on. I don't want my car's automation to remove this life-saving decision from my hands.

I can see it now, avoiding by turning right, and putting me right in the path (driver's door) of death.

Rather than continue to turn out more gimmicks (which do nothing but interfere with natural selection), how about a car that can get 100 Miles to a gallon. Nissan: please go back and work on this one for a while and leave us alone.

No. Just No.

[epp_b]

There are just too many unpredictable situations that a computer just can't calculate.

Just like ABS, traction control, stability control ... they're all just ways of allowing drivers to become stupid, lazy and less involved. If it were up to me, automatic transmissions would be illegal (except for special cases like disabled drivers, etc.), I really think it caused drivers to pay too little attention.

We need to stop trying to mitigate stupid drivers and just get rid of stupid drivers, either by improving the training regimens or getting them off the road and providing them with viable transportation alternatives.

Re:No. Just No.

[pclminion]

The less involved the drivers are the safer the system is. The system is safest when the driver cannot control the car at all.

Re:No. Just No.

[lordsilence]

While this is true. There are quite a few situations that can be calculated. For example, a driver losing falling unconscious / falling asleep would be a perfect example of a system that could kick in and safely slow down the speed of the car. I'm not saying that a system such as this should be trigger-happy, but simply kick in as a last resort.

Re:No. Just No.

[Barbarian]

The less involved the drivers are the safer the system is. The system is safest when the driver cannot control the car at all.

The Airbus school of thought.

You can excees the design of the engines ib a Boeing aircraft in reverse thrust to stop it from overrruning the runway in emegency conditions. The engines will be damaged and need overhaul but you can avoid becoming a giant fireball. In an Airbus, you cannot.

"If only the crew kept their sinning hands off of the controls"

Good luck with that.

[Tastecicles]

You'll never get it in your carhole.

Re:Sheesh! Programmers!, What do *they* know

[jrmech]

I believe it's pronounced "Mechanical Engineer" not "Programmer"...

What could possibly go wrong?

[ntropia]

Let's see, if we trust the summary to be a real summary of the article (no, it wasnt' enough to trigger my interest in reading TFA), there are only few [sarcasm] points where the choice of parameters for the training of this feature could go wrong:

radar and laser scanners: false positives? sensitivity? distance cut-off? Am I going off-road because of a restaurant at the end of the road?

front-mounted camera: image recognition accuracy? influence of light condition/weather?

alert the driver: how is the driver going to react to a very friendly red-blinking "Imminent death" widget in his dashboard, or a soft-voice "We'reAllGoingToDie" reminder? AirFrance disaster anyone?

if the driver cannot immediately turn: what's the timeout? why he can't turn? Maybe there are obstacles on the sides the front-mounted camera can't see? Just sayin'...

Oh, and let's not forget the most important one!

A new computer system automatically steers the car

That's possibly the worst that can happen when there's a driver trying to do... well... the driver.

Human judgment is needed in picking a direction

[SuperKendall]

First, it's dangerous to drive into any obstacle. You do you know that box/balloon/etc. doesn't have a bowling ball in it?

Given he was describing it as "drifting", not rolling, across lanes I think it would be pretty obvious which you were dealing with.

Second, why do you assume the system would be any worse at choosing among the options -- left/right/straight -- than a human would?

Because it's very hard to have enough sensors and computing power to make that choice.

Ditch or nice smooth road off to the right? An easy choice for the computer, the road. The human driver? Probably not going to pick the smooth road where he can see just cresting the hill a semi truck heading the opposite way at 70MPH.

Or possibly either are really bad and you should just hit whatever it is you are going to hit because that is the best possible option. How long before such a system can decide that is the right thing to do?

Re:Human judgment is needed in picking a direction

[AmiMoJo]

The logic is simple. You slam the break on, just like a human driver would, unless you are absolutely 100% sure that swerving is safe. That is why the computer has multiple cameras and radar to check, and since it can see in all directions at once it can be even more sure than a human driver.

Re:No thanks.

[green1]

The problem is, these won't be optional.
How easy is it to find a car now without ABS? or even traction control? both of these systems cause far more problems than they solve with even a half way competent person behind the wheel, and yet you can't buy a vehicle without them. This will just be one more mandatory expensive system that you MUST pay for on every new vehicle, and if it causes a crash, it will be considered your fault just as it is today if your ABS or traction control do the same.

What is really needed in the automotive safety world is to make sure drivers are half way competent before we let them out on public streets. Unfortunately that sounds too much like personal responsibility and how dare you deny someone the RIGHT to drive just because they don't know how!

Possible scenario

[DodgeRules]

The driver of the car in front of you jams on his brakes. The road is wet and your car can't stop in time. There is a truck to the left so your brand new intelligent car decides to swerve to the right because there is only a small object there and won't cause as much damage. Too bad for the student walking home from school.

This idea, while the concept has good intentions, just sounds like a disaster waiting to happen with a huge lawsuit for an ending.

Re:Possible scenario

[jamesh]

The driver of the car in front of you jams on his brakes. The road is wet and your car can't stop in time. There is a truck to the left so your brand new intelligent car decides to swerve to the right because there is only a small object there and won't cause as much damage. Too bad for the student walking home from school.

This idea, while the concept has good intentions, just sounds like a disaster waiting to happen with a huge lawsuit for an ending.

Probable scenario. The driver of the car in front of you jams on his brakes. The road is wet and your car can't stop in time. There is a truck to the left so your brand new intelligent car decides to slam on the brakes and hope for the best, because there's a small object to the right that might be a student walking home from school.

Seriously, it's not rocket science. Your car isn't going to just "aim for the softest target". And in any case, if the car in front of you can stop in time then your computer assisted braking can stop you in time, assuming your vehicle has similar stopping capacity (tyres not too worn, suspension in good order, not overloaded, etc). With your reaction time replaced by the computers reaction time you are in a way better situation than you would have been without this technology.

Re:Possible scenario

[rastos1]

The driver of the car in front of you jams on his brakes. The road is wet and your car can't stop in time. There is a truck to the left so your brand new intelligent car decides to swerve to the right because there is only a small object there and won't cause as much damage. Too bad for the student walking home from school.

He was eaten by a grue?

Re:Possible scenario

[rgbrenner]

How does it know the object "won't cause as much damage"? That small object could be a concrete block. It all looks the same to the computer.

So what gives you the idea that it would swerve into what could be a concrete block?

Re:Possible scenario

[AmiMoJo]

Obviously it won't swerve if there is anything at all it could hit. It doesn't decide what thing is the least bad thing to crash into it just avoids one specific collision if there is absolutely no chance of causing any other. If there is even the slightest doubt it won't activate.

Re:Possible scenario

[quacking duck]

The driver of the car in front of you jams on his brakes. The road is wet and your car can't stop in time. There is a truck to the left so your brand new intelligent car decides to swerve to the right because there is only a small object there and won't cause as much damage. Too bad for the student walking home from school.

I wonder if the intelligent car might not enforce the so-called 2-second rule. The counter to this is that other drivers will just cut in, making the car fall back again and again... but maybe the counter to *that* is that it's only engaged when it detects adverse driving conditions (rain, snow, freezing rain, dust storm, etc). Assuming the laser/radar/magic system doesn't see hard rain as a solid enough obstacle and prevent you driving at all.

In the year 5555...

[ElitistWhiner]

http://www.youtube.com/watch?feature=player_detailpage&v=WhNM2K8cmU8

In the year 5555
Your arms are hanging limp at your sides
Your legs got nothing to do
Some machine is doing that for you

lyrics: http://www.lyricsmode.com/lyrics/z/zager_and_evans/#share

Re:No thanks.

[arth1]

How easy is it to find a car now without ABS? or even traction control? both of these systems cause far more problems than they solve with even a half way competent person behind the wheel, and yet you can't buy a vehicle without them.

The problem is a dearth of halfway competent people behind the wheel. Even if only a tenth of all drivers were incompetent, mandating things like ABS and traction control will have a positive impact on accidents. And reality is that most drivers would lose the cool and slam on the brakes in a dangerous situation.

IMO, we need two levels of driver's license. A higher level that requires true training, and is much harder to obtain, and which brings privileges like being allowed to disable certain "safety" features when they have shown they can do it manually as well as or better than the expert systems, or exceed speed limits on certain roads like divided highways.

Re:No thanks.

[swalve]

This has been proven wrong time after time. Your halfway competent driver is just a lucky one. The car can react far faster than you can.

They're trying to idiot-proof driving

[Beeftopia]

There's actually not a lot of full blown idiots out there. But the ones who are out there leave a trail of death and destruction. To the tune of about 30,000 dead a year, plus some multiple significantly injured.

As someone who has well over 10,000 hours behind the wheel, and a good driving record, I don't think the roads are going to be made safer with this system on my vehicle. I already keep my eyes forward, maintain sufficient following distance and slow down in bad weather. However, for a teenager, an elderly person, a drunk driver, or just someone who is stupid and gotten hold of a car, I think something like this will be very helpful in reducing roadway fatalities and injuries. Disabling the system should require the factory key and a passcode.

Car automation

[NetGyver]

I wonder how the system would handle animals on the road. I don't know exactly how many animals I've hit since I've been driving, but I'd bet it's been dozens. 3 deer, all on the highway, some raccoon, some squirrels, some opossum. I was in a situation where I literally had to choose between swerving and wrecking in to the cars beside me on the freeway or hitting the deer. Needless to say, if its an animal, it's game. But, OTOH, computers are faster than humans, and sometimes things happen so fast that you don't get time to think.

In the split second between imminent danger and reaction, the speed of your vehicle, it's trajectory, and where your likely to have impact...these are all things a computer can calculate nearly instantly. I am not saying Yay for driver-less cars, i'm merely saying that If your going to crash, and with so many variables associated with crashes, at least if it's computer controlled, your risk is calculated, instead of solely relying on random reactionary impulses.

There is always that scary feeling like you lost control. Humans like to give control up for mundane and repetitive tasks, However, this is literally putting your life in the hands of a machine.

I'm more in favor of safety warnings that get noticed by the computer and escalated to the driver. For example, if your switching lanes and a car is in your blind spot, and you put your signal on, the computer warns you there is a car there in the blind spot. That's awesome. Or if standards were in place cars would talk to each other in such a way that Jimmy's BMW could bug the shit of of him for being a dick and cutting me off and getting too close while doing it...that's awesome too.
Or if i'm getting on the freeway and my Jeep knows that the guy coming down the road is going too fast and chances are high that if I try to merge on, he'll hit me in the rear. Or, (and this one is great) if your trying to get out on a busy road, the system could let you know when it is truly safe to pull out on to the main road, based on the speeds of the oncoming cars.

There are all kinds of things a computer system can do to make us safer drivers, without outright taking full control of the situation.

Re:Car automation

[L4t3r4lu5]

I was in a situation where I literally had to choose between swerving and wrecking in to the cars beside me on the freeway or hitting the deer. Needless to say, if its an animal, it's game.

Actually in this instance it's venison.

Re:Car automation

[Jeremi]

Needless to say, if its an animal, it's game.

I'm sure there will be a "redneck mode", where any time the sensors detect an animal on the road, they direct the car to swerve toward it. (the robotic arms to automatically retrieve, skin, and dress the carcass will be a costly extra)

ok...

[Charliemopps]

and when the car automatically "Avoids" some obstacle, real or not... and plows into a 4yr old on a tricycle... Who's liable? This seems like a horrific idea that I can't believe Nissans lawyers would allow on the road.

50/50

[epyT-R]

This could just as easily CAUSE an accident. Really, computers should be ASSISTING the driver, not making decisions for him.

The trouble with shared control

[Animats]

Fully automatic driving is likely to work better than trying to steer out of trouble just before a collision, after the driver has already botched it. A full auto system, which doesn't have an inattention problem, is more likely to see the problem early. What you'll see with a full auto system is a tendency to slow down in ambiguous situations until the picture becomes clearer. This will probably happen more often than with human drivers, but won't add noticeably to trip times.

And how well does it make moral decisions?

[ezakimak]

A driver may well choose property damage over bodily harm--eg crash into another vehicle rather than hit a pedestrian that radar may not even detect.
I cannot believe this will sell.

A proposal

[gman003]

I have no problem with handing more control of a car to a computer, rather than the person. People are stupid, after all - every accident I've seen or been involved in, some stupid person was at fault, and that *includes* the time I wrecked my own car.

I just have one request. Put a big, easily-slapped button, preferably red, on it that completely overrides everything. EVERYTHING. Collision avoidance. Lane-follow. Auto-braking. Fucking cruise control. I should be able to kamikaze the car into a building as long as I'm pressing that button. And there should be no cutoff or limit or penalty - if I duct-tape it down, it should disable everything just as I asked. And the only penalty should be that, if I use that button and cause a crash, it counts as deliberate, not accidental. No "you pushed that button once, your insurance just tripled", no "you pushed that button for no reason on our list of acceptable reasons, you get a fine".

If you do that, I'd even accept more restrictive controls. You could put in a speed cap that cuts the accelerator off when it hits the highest speed limit in the country (85mph, I think). Or something to prevent you from changing lanes without using your turn signals. Or something that cuts the engine if you pull out your phone.

Killer cars

[Anonymous Coward]

Driving country highways at night I frequently flatten animals that are in the highway for one reason or another. Raccoons, opossums, etc. It sucks to hit them, but what's Nissan planning to do - toss me off the highway at 60mph instead? No thanks, I'd rather plow ahead and deal with any damage (usually none) that it may cause.

Even with larger animals like deer it's best to hit the brakes and keep the wheel straight than it is to try swerving at high speeds. At 60mph a swerve will most likely result in you flipping the car multiple times, whereas hitting a big animal straight on during a hard deceleration may not do anything more than give you some front-end damage.

Machines can't reason, they can't consider context, and they can't predict severity of consequence of one decision over another. There is no way that this move by Nissan doesn't have some very bad outcomes.

ssan Motor has deep pockets

[Joe_Dragon]

Nissan Motor has deep pockets and if this fails and some get's hurt or dies not only should a big law suit there will need to be a government investigation as well.

and not they can't hide by subbing work out and trying to hide under a EULA

Re:No thanks.

[compro01]

with even a half way competent person behind the whee

Let me know if you find one.

well ... no

[Osgeld]

If I know in enough time that someone is going to jump out at me to let go of the wheel, there are plenty of other corrective actions I would perform ... if the wheel suddenly jerks the first think in my head is that I blew a tire or hit something or both and try to correct it before my car goes into the ditch, not let go and shrug... possibly causing a even bigger problem the moment this thing lets up and my muscles are still fighting a servo.

how about people just dont haul ass on a street where people might walk out? dude is fucking flying for a simulated road where this situation would happen.

Bad idea...

[hachre]

If my car were to drive fully on it's own, then yes - automatic collision prevention is a must - but a car that tries to make a guess about what I'm going to do and takes control away from me in a very shitty situation is definitely a horrible idea.

Pilot by Committee

[Roger W Moore]

Just like ABS, traction control, stability control ... they're all just ways of allowing drivers to become stupid, lazy and less involved.

Not just less involved but less in control. They either need to go the whole way and have properly driverless cars where the computer is responsible for everything or have the driver responsible for everything which means that a computer cannot override the drivers control. Having both driver and computer each partially in control would be like having an aircraft piloted by committee.

Re:No thanks.

[hankwang]

"we need two levels of driver's license. A higher level ... which brings privileges like ... exceed speed limits on certain roads like divided highways."

Surely you believe that you will pass the test, since you know that you are an above-average driver?

  http://en.wikipedia.org/wiki/Illusory_superiority#Driving_ability

Old maids tales

[dutchwhizzman]

Yes, the first generations of traction control systems were rather crude and didn't work well on ice. However, some manufacturers actually evolve their product and modern TC systems are doing a whole lot better than the ones that caused all the urban myths. That doesn't mean that every car you buy now has a capable TCS on it, but if you steer away from "We sold this model for the last 100 years, why change?" or "Our car is cheaper because we copy old Euro tech" style manufacturers, you will probably find capable electronics that actually help prevent accidents. However, there is no recipe against a driver that chooses to drive way too fast under circumstances that will most likely cause ice on the roads. Once you go 50 mph in a certain direction and hit ice, there is no amount of electronics or steering that will stop you sliding straight ahead in the direction you were going. It often takes a tree, house, rock, car or other large object to stop you, or you'll roll over once you gain grip again or hit a ditch. Physics can be a bitch sometimes and there will always be people that fail to realize that in time.

This new technology isn't about traction control, it's about not hitting that pothole or lost cargo on the road. This will mean that your car will suddenly swerve hard, slamming the steering wheel out of your hands, breaking your thumb or fingers and making you spill your hot coffee on your lap. People will blame that on their car, but they fail to realize that you shouldn't be holding the wheel with just one hand, or with your thumbs hooked, or drinking hot coffee while driving. It's the same as with driving under icy conditions; these mechanisms are put on cars to help good drivers deal with situations the human brain can't cope with, not to substitute the driver. If you want that, go talk to Google.

Re:No thanks.

[adolf]

Nay, three levels:

Fully-automated self-driving car license (trivial to get, and why not? all you have to do is keep it maintained and make sure the sensors are free of obfuscation, the rest is just a bus ride)

Semi-automated car license (similar to our current driver's license requirements) for cars that can avoid things on their own, but still need operated primarily by a human

Advanced car license (requiring comprehensive training and testing, but allowing for certain merits in consideration).

This isn't all that dissimilar from the different classes of CDL currently offered in the States.

(And I suggest this even though I'd be "forced" to undergo more training and testing to get an advanced license to keep driving my current cars. That's where the consideration comes in...)

Re:No thanks.

[madsdyd]

Interesstingly, about 80% of drivers believe they are better drivers than average.

So does everyone else

[SmallFurryCreature]

And everyones penis is 10 inches too. Even the girls. Or maybe that is just the bars I visit. Good coffee though.

Re:God, I love Japanese scientists/engineers

[hcs_$reboot]

God, I love Japanese scientists/engineers

Nissan is managed by Renault/France/Carlos Ghosn.

SKYNET Traction Control? Really?

[triffid_98]

First off, The most effective way of handling snow and ice is a light car with thin snow tires. That's just how physics works. Grip/Weight.

Second, Traction Control usually means throttle control. If tires start to slip the ECU (Engine Control Unit) reduces engine power. In AWD cases it may shift power to different wheels or in more advanced systems it may modulate specific wheel braking but that's a best case scenario.

The bottom line here is whether your ECU is a better driver than you are. I'd like to believe it's not, but most people have no idea what to do if their vehicle starts to skid so this is probably a benefit, provided you (and those pesky Ambulance Drivers) can easily disable it. Everyone should know how to handle their car in a skid but it's not something you'll get in your typical drivers-Ed course.

Re:No thanks.

[arth1]

Surely you believe that you will pass the test, since you know that you are an above-average driver?

No, where did I say that? To be of value, the second level license should be so hard to qualify for that most people won't qualify, and certainly none without specialized training.

While I do have track training, and have demonstrated to an instructor that I consistently can stop a car faster without ABS on slick tracks, and do avoidance manoeuvres, that doesn't qualify me. It might make parts of it easier, but there are lots of things that I have no hands on experience with. Could I bring a front wheel drive out of a spin? No idea - probably not. Am I the best parallel parker in the world? Quite certainly not. And my reaction times are not those of an 18 year old, even though that's probably somewhat compensated for with the brain having developed filters prioritizing traffic events.

Do I think I could pass such an evaluation and exam though? I believe I'd have a good chance, if investing the time and money. Just as I believe I'd have a good chance of getting a master's degree in physics too, if I put effort into it. But I don't think most people would put in the effort for either. And I don't know for certain that I'd pass.

Re:Waiting for the first report...

[realityimpaired]

My car has a speed limiter in it... it kicks in at 185km/h (112.5mph), mandated by the Japanese government as it was manufactured there. If you need to go that fast to pass somebody, maybe you shouldn't be passing them.

More likely scenario, you hit the rev limiter, in which case you change gears, you don't bitch at the limiter.

If the driver wants to run over the butt rapist...

[Tristao]

Can Nissan be brought up on charges of sodomy accessory (not to confuse with the accessories one might use upon going route 3)?

Re:No thanks.

[coofercat]

We already have this - it's called the Advanced Driver test (see http://www.iam.org.uk/). You can get insurance discounts if you've passed one. I'm told one of the tests they make you do in the exam is to drive as fast as you can down a windy single track road. The thing is, it's a test of your safe driving, not so much how fast you can go.

Yes. Just Yes.

[dabadab]

Just like ABS, traction control, stability control ... they're all just ways of allowing drivers to become stupid, lazy and less involved.

And they are also great ways of avoiding accidents. I prefer being "lazy and less involved" (I fail to see how "stupid" comes into the equation) to being dead or crippled, thankyouverymuch.

We need to stop trying to mitigate stupid drivers and just get rid of stupid drivers

Well, if history is any indication, getting rid of stupid drivers Just Does Not Work(TM) *, while all the stuff to mitigate the dangers does, just look at the steady decline of fatalities per mile travelled

*: furthermore, you personally might be offended by your not being allowed to drive any more, unless you meant to "get rid of all the stupid drivers, but me", of course

I've got a cheaper solution

[Dcnjoe60]

I've got a cheaper solution. Instead of adding technology to make the car steer itself because you don't have time to react to the obstacle in the road - lower (and enforce) speed limits. 60mph instead of 70 give you 15% more reaction time and 15% more distance to react. If this is truly about safety, slower speeds are the answer. However, if it is about trying to make it safe to be able to drive ultra-fast then the technology is probably a good idea. Then again, lower speed also has the benefit of reducing demand for gas and oil, less wear and tear to the vehicle and pavement and a number of other benefits.

Re:No thanks.

[arth1]

We already have this - it's called the Advanced Driver test (see http://www.iam.org.uk/ [iam.org.uk]). You can get insurance discounts if you've passed one. I'm told one of the tests they make you do in the exam is to drive as fast as you can down a windy single track road. The thing is, it's a test of your safe driving, not so much how fast you can go.

We need this in the US too. Preferably combined with reasonable and not-too-crippling limitations for those who don't pass the test, like not being allowed to drive without ABS or EPS.

So basically, the machine overrules you.

[Revotron]

The system first alerts the driver to turn in a certain direction. If the driver cannot immediately turn in that direction, the system takes over the steering to help avoid a collision.

Either this system is incorrectly designed or the article is incorrectly summarized. What this article is saying is, if something runs out into the middle of the road, and the computer recognizes both sides of the road as clear and randomly decides I should turn left, but I veer right based on my own thought process, habits, etc, the computer will take over and turn the wheel back to the left as I'm already turning right? Thereby countering my original veering and possibly putting me right back on course to ram head-on into said obstacle?

What a grand idea.

No it can't

[SuperKendall]

That is why the computer has multiple cameras and radar to check, and since it can see in all directions at once it can be even more sure than a human driver.

But it cannot see very far because of blocks from cars in front and behind. If the driver was paying attention they know what traffic is to the right and left for a much longer way than the computer reasonably can with technology we might have for the next twenty years.

The computer can make great choices about stuff right around it, but not about choosing which way to deviate from a lane.

Re:No it can't

[AmiMoJo]

Right, but if the computer can't see because the view is blocked then it won't swerve. It only does it if it can be absolutely sure of not causing an accident.

Attacking the Driver

[Anonymous Coward]

If a vehicle was going to take steering out of the hands of the driver, I'd hope that it would be done such that the steering wheel is not involved. A steer-by-wire system offers a mechanism for doing just that. With such a system, the car would swerve and correct, with the driver continuing to hold onto the steering wheel all the while. It would be insane to yank the steering wheel from the driver's grasp unless the car was going to continue to self-drive long enough to allow the driver to recover from whatever mayhem might ensue in the driver area.

There may be unwise ways of holding the steering wheel, but I don't see a reason to attack the driver simply because the car designers have decided that the car should have a "self-steer.event". I'd hate to be a diligent driver with a firm grip on the wheel during one of those events.

Mod parent up.

[bussdriver]

Wish I had points. people don't like the truth.
The costs on roads for higher speeds are amazingly HIGH. Especially when you kill your infrastructure and move everything by heavy trucks. F=ma.

Re:No thanks.

[lewiscr]

I could see about 80% of drivers being better than median. It's hard to be a really good driver, but there no bottom on bad.

What about Bridges? my 1st thought

[bussdriver]

Open space on the left and the right... the right side is preferable (no oncoming traffic) so turn right... and OFF THE BRIDGE!

Ok so bidges have side rails that are usually solid (but may not be) but what about most bridges which lack side rails before and after the bridge? There are really steep hills leading up to many bridges.

As you launch off the side hill, the car sees the ground quickly approaching you don't respond quick enough due to the shock of being airborne so you hit the ground with your wheels turned right; breaking your steering mechanism and causing the landing cart to veer right and tumble down hill sideways.

Then it is pointless

[SuperKendall]

Right, but if the computer can't see because the view is blocked then it won't swerve.

So then it would never swerve because it cannot see far enough. In the end then the system is utterly pointless if it will only activate in cases where it can see far enough it would never need to activate. The only time you need to swerve is when you are close enough to something the computer cannot see.

Re:Waiting for the first report...

[ThatsMyNick]

May be, he is passing someone on the autobahn?

Re:No thanks.

[green1]

If we actually had testing standards before issuing licenses, most people would be able to qualify, with some instruction. Unfortunately current road tests are a joke.
I remember taking my test and realizing they cared more if you could read signs and park straight then if you could actually drive. No skid training was required, no tight space manoeuvring, in fact no emergency situations were tested at all. They cared much more that you could parallel park (in a spot the length of 3 city buses) than they did if you were an actual safe driver.
While I'll admit that not everyone could pass such a test, I bet the majority could with a bit of training. But as it is, the training isn't required, so very few people get any.
If you combine that with measures that stop people who have been proven at fault in multiple collisions from continuing to drive without further training, you'd end up with much safer roads.
The solution isn't better features on the cars, it's better training and more accountability for the drivers.

Re:No thanks.

[compro01]

That would work great if practically every city in the country wasn't built around the assumption that everyone can drive.

Re:No thanks.

[green1]

Well, if driving is an inalienable right that everyone should have regardless of knowing how, we should do away with the farce of driving licenses. If we believe that licensing is important to ensure safe drivers, we should actually do it.
I live in a city that is positively AWFUL for people without a car, and yet I lived that way for years without one. It is doable, and if you want to drive, you'll learn how. I don't think these standards would make most people unable to drive, it would just force most people to actually learn how. sure a few people would be disqualified, but do we really want those people driving?

Could be good.

[ebvwfbw]

Bad thing about anti-lock brakes is that people are depending on them more and driving more aggressively. I hope the same thing doesn't happen here.