Slashdot Mirror
California DMV Told Google Cars Still Need Steering Wheels
cartechboy writes Google showed us what it feels is the car of the future. It drives itself, it doesn't have a gas or brake pedal, and there's no steering wheel. But that last one might be an issue. Back in May California's Department of Motor Vehicles published safety guidelines aimed at manufacturers of self-driving vehicles. After seeing Google's self-driving car vision, the California DMV has told the company it needs to add all those things back to their traditional locations so that occupants can take "immediate physical control" of the vehicle if necessary. Don't for a second think this is a major setback for Google, as the prototypes unveiled weren't even close to production ready. While the DMV may loosen some of these restrictions in the future as well all become more comfortable with the idea of self-driving vehicles, there's no question when it comes down to the safety of those on the road.
In fact, it would not surprise at all if the brake itself is NEVER removed. I can easily foresee a situation where these vehicles are used to transport unwilling people, or simply undergo a malfunction and the occupant will always want the ability to stop the device.
But I can see the steering wheel and accelerator going away completely - don't want to let untrained people having the ability to make things worse.
excitingthingstodo.blogspot.com
Personally I'd like to get a car that has controls similar to a jet fighter - or even more basic if it's all drive-by-wire anyway. Gimme a throttle lever in one hand, and a twist stick for proportional steering in the other - or combine them. More display room and less clutter of a wheel.
Never have a philosophy which supports a lack of courage
Is the human driver going to be "responsable" for failing to take action if his autonomous car goes haywire and causes an accident ? I mean, if the car besides the AI is going to have a steering wheel, gas pedal, brake pedal, clutch pedal; what's the advantage for the human passenger ? You'll end up babysitting the AI for the whole drive. No thanks, just use a normal car.
What if I happen to be a much worse driver than the car's software and in "wresting control" away from the car I inadvertently cause an accident that the software could have avoided (or was in the process of doing)?
Have they not seen "I, Robot" (2004)? Of course you need a manual override.
.. So that real horses can take "immediate physical traction" of the vehicle if necessary.
Any car that allows the driver to take "immediate physical control" makes the roads unsafer for all. The safest roads will be when ALL cars are autonomous. Having humans in the mix will just ruin all the gains that autonomous cars provide. Can a human wirelessly communicate with a car 5 miles ahead to know of a road condition and adjust it's speed in tandem with all the other cars in between to mitigate any and all danger in advance? Can a human react in sub-millisecond time to avoid obstacles thrown in their way. No and no.
Eventually it will become clear that human drivers are the weakest link.
Consider how many people die on the roads every year in the United States alone, the biggest factor is humans.
Then, consider how many people die every year due to firearms in the United States alone and look how CA reacts and tries to limit access to firearms, through laws and technology. Basically, "remove the problem."
Shouldn't CA be pushing hard for driverless vehicles? Removing human error from the equation would save countless lives.
Think of the children.
Which is: 2 out of three of those are already physically disconnected in a modern automobile, and the third they're working really hard to do the same with.
If California's DMV wants to bitch about google removing all those 'unnecessary' features, then it should look really hard at what it's allowing in it's HUMAN OPERATED vehicles.
Allowing a computer to adjust throttle or braking response is one thing. But physically disconnecting the pedals from the devices they're meant to operate means in the event of an electronics failure there's no guarantee you'll have control of braking, throttle, or even steering anyway.
Food for thought!
I think it would be nice to still have a steering wheel. Not that I doubt the safety and precision of the vehicles, but it is always good to have manual control just in case some sort of freak accident or whatever occurs. For example planes can fly themselves as well but we still have manual controls for those just in case, even though the situations that they may be needed in are probably extremely rare.
That's basically all is needed for an autonomous vehicle.
They still leave the operators open to liability, and he/she has to pay attention. The vehicle is not autonomous until then. Babysitting the damn thing is not part of the deal.
“He’s not deformed, he’s just drunk!”
If you keep the brake, steering wheel and other human usable control interfaces, then you can still hold the human liable. I would love to have a self-driving car but I'm not going to get one if I am liable for accidents as I won't be driving, so how could it possibly be my fault?
That means I'll just keep driving to the best of my abilities and hopefully they can keep adding in small corrective things to assist in the driving. I love my cruise control and I hear that some of the newer and fancier cars let you toggle a setting that basically will pace the car in front of you. How awesome is that?
All they really need is a Big Red Button to just power off the car and brake as fast as possible - auto ABS, maybe?
As far as steering out of a spin....you got me there. Then again, how many folks know how to do that anyway?
I agree that an automated car will need a steering wheel in the immediate future. Once their track record has been proven and people are comfortable with them, however, cars will gradually lose manual controls. We'll likely be telling our grandkids with stories of hundreds of non-automated cars screaming down the highway piloted by fallible humans. Of course, they'll just roll their eyes at us, make an "uphill both ways in the snow" comment, and tell their RobotCar to take them to the mall.
My sci-fi novel, Ghost Thief, is now available from Amazon.com.
They still want permits and license plates to be a thing, so they don't want you do be able to just hop in a car and be drove around as you want.
I would sooner trust my life to a self driving car, than to one with a meatbag at the wheel.
If there's one lesson I learned from Star Trek it's that you always, ALWAYS, include a manual override.
The transition time from the computer giving up to the user having to take control is always going to mean this is impossible.
If you're reading the newspaper, you are not going to be able to transition to operating the vehicle in the event the computer gives up and says it's all up to you.
I've been saying for a while, that a driverless car needs to be 100% hands off for the people in the car, or serves no value at all other than as a gimmick.
I will believe driverless cars are ready for prime time when I can stumble out of a pub, crawl into the back seat and tell the car to take me home. Anything less than that is a giant failure of automation waiting to happen, and a convenient way of dodging liability by pretending that users are expected to be in control of the car even while the AI is driving.
As long as there is a pretense of handing back to the driver in even of an emergency, this is a glorified cruise control, and I'll bloody well drive myself.
If I'm ultimately responsible for the vehicle, I'll stay in control of the vehicle. Because if there's a 10 second lag between when the computer throws up its hands and says "I have no idea" and when the user is actually aware enough and in control, that is the window where Really Bad Things will happen.
Lost at C:>. Found at C.
In fact, it would not surprise at all if the brake itself is NEVER removed.
That's the current situation with driverless trains/subs:
No cockpit in the front of the wagon, but you still have "emergency brakes" lever everywhere.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
So just how do these driverless cars deal with heavy fog, rain, sleet, snow, icy roads, etc? I'm all for them and can't wait for them to have full penetration and that's all we use, but I've yet to see anything about how they do in bad weather. Until that is proven, not only should steering wheels still be there, along with brake pedals, gas, etc, but a real, licensed driver, still paying attention to the road, should still be required to sit in the driver's seat.
Is this requirement based on science or an irrational fear of computers?
I'm glad they will include a manual override.
Yes, driverless cars are likely to be safer than human driven cars. However, what will they do when an emergency occurs and the road is diverted in new and exciting ways that not only are typically illegal, but are mildly dangerous? Simple, they'll shut down. No steering wheel? Now you just caused a traffic jam on a major artery.
The busiest highway in North America has been diverted to the point of people driving the "wrong way" before (of course with plenty of officers present to help traffic move). Did you know that a driverless car that shuts down when it gets to that "Uhhh... I don't understand hand signals? You want me to drive over the median to the wrong side of the road???" point would be fined an enormous amount of money per hour?
As long as there is some ultimate fail-safe button that brings the vehicle to a controlled stop why would you require a steering wheel/brake/gas? Maybe for the sake of simplicity put it where the brake would usually be, make it pressure sensitive so the vehicle knows how to respond (slow down or slam on the brakes). Under normal operation a person isn't going to be paying enough attention to take immediate steering control & a gas pedal isn't really necessary for any real world road safety situations.
I wonder if a human could withstand the deceleration caused by a boat anchor and cable. In case of emergency break glass by tossing boat anchor out the nearest window while keeping clear of cable.
If this is an infancy thing, fine, but I don't think we need to reach Ultimate Endgame for these things to be ride'able (NOT "operable") by a child, or even unmanned for deliveries. It'll be the better choice long before we reach superduper polished and perfected.
Mind, I still expect them to come with some tucked away form of control access, even a clunky digital-only one. There's an endless number of possible edge cases that can't be scripted.
I assume code will mostly fail-safe into "Stop the car." for unexpected encounters, so that someone (incl remotely, ie support center or the owner) can help manually guide the derping sensors back to the clean, plainly marked road and resume automated driving, or summon a tow for really locked up/physical damage situations.
90% of accidents (or more, depending on the study) are due to human error. So the DMV insistence on putting the humans back into the drivers seat is actually counterproductive. "there's no question when it comes down to the safety of those on the road." ... the question is are the other humans on the road more or less safe with the google vehicle operators able to override the computer?
While I'm not interested in being an early adopter of this or most automotive technologies, there are lots of questions when it comes to safety. It is a pity that government hardly ever uses science or logic in the decision making.
After you've operated them safely for 20 years. I'm all for progress, but let's not skip over the part where there is a trial period while society adapts to changes.
Can a human wirelessly communicate with a car 5 miles ahead to know of a road condition and adjust it's speed in tandem with all the other cars in between to mitigate any and all danger in advance?
Do not assume that source of wireless coordination is always 100% trusty.
The wireless coordination information might be hostile origin. i.e.: some idiot with a hacked emitter that systematically ask all the other cars to slow down and move aside to let him go through. In theory such a function has practical uses (ambulances, for example), in practice such function WILL GET abused (idiot wanting to arrive faster, or a criminal trying to run away through heavy traffic).
Can a human react in sub-millisecond time to avoid obstacles thrown in their way.
Yup, that's what I consider as the main reason why we should have robotic drive.
Except for the occasional false positive, the current collision-avoidance systems that are already street-legal nowadays and that are already travelling in some cars around us are already much better than humans in reacting in case of emergency.
The only drawbacks currently are false positive[1].
But even in that situation, most of the false positive are safe.
It just causes the cars to slow down or stop when that should not be needed.
---
Example on our car:
- auto-cruise control which chooses the wrong taget: with our car, a large truck that is almost as large as the lane can be mis-targeted and our car slows down to yeild, even if the truck is actually in a different lane and we're not actually on a collision course with it if we stay in the middle of our current lane.
- mis-identified target: the current logic inside the car is: "if there's an object on the way and the car is on a trajectory intersecting it, then hit the breaks (unless overriden by the driver)". The car has no concepts of *what* the object is, and might break on useless occasion. Nearby automatic RFID-based tool booth are non-stop drive through: you don't need top stop, just drive through at a slow steady pace. The RFID transponder will beep in advance to alert you that the transaction with the booth was successful and the barrier will open shortly, you know that barrier will open shortly/is opening shortly and you don't need to brake. But the car only sees an object that is still currently inside your lane (it's not able to notice that the object is moving vertically and that by the time you reach it will be safely away) and will auto-brake unless you keep your feet on the gas pedal.
- very simplified hit-box: the car's hitbox is exactly that: a box. the car will panick and hit the brakes if you try to park under a low hanging balcony. You see that there's enough room under the balcony for the car's engine compartment to go there, but the car will react as if it was a solid wall and break if your foot is on the brake instead of the gaz pedal (which will be the case during slow manoeuvres).
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
that this may well CAUSE more accidents that it prevents? also, that delaying the roll out of self-driving cars is almost certainly a net negative from an accident/injury/fatality standpoint?
false dichotomy - my personal favorite of all the fallacies...
There is an error in your sampling. Given that 100% of accidents have human drivers involved, it's not a huge surprise that you find a large number of causes point to human drivers, which are the most complex piece of the system.
Substitute a different piece, an automated driver program, then that becomes the most complex piece of the car, maybe not as complex a human, but it's where we're likely to see the most failures when we analyze crashes in the future.
Lets us say that in a ideal world, the requirement for a steering wheel, brake and accelerator was *not* required.
Then....
I can get driven home drunk, as I am not driving(or ever expected too)
A 14 year old *could* get driven to the after school after school practice, and then driven home. (no need for a license)
The self driving car is like a subway, or bus. As a rider, my expectations will be in line with public transit, yet, I will have
more options as to where and when I go.
The requirement for all those other controls, still means it is not self driving.
Of course I understand all the reasons being cited as to why they are needed. I just disagree, I believe that consumers will be able to
decide if the real self driving car is safe and reliable or not. We still have lawyers, and that threat is the single most compelling reason
that any manufacturer will make them viable.
When cars first came out, some cities made stupid laws such as the car had to flag man walk in front of the car(safety, etc)
If a driverless car has no manual means of steering, and if it broke down and you had to push it, how could you control it?
The laws will be rewritten once this gets closer to being a real thing. Google can continue to do what it wants on its test tracks.
So that real horses can take "immediate physical traction" of the vehicle if necessary.
You have no idea how punishing the roads were in the early days of the automobile, how often cars broke down or became hopelessly mired in mud or snow. In rural states, the horse was still in the towing business as late as 1940.
In 1919 Lt Col Eisenhower, yes the later Supreme Allied Commander of WW2 and the 1950s President of the US, led a convoy of 24 vehicles from the east coast to the west coast. 9 vehicles were lost, 21 men were injured and unable to continue.
"... stumble out of a pub..."
Like the inebriated gentleman in San Francisco of many years ago? He stumbled out of a pub, crawled into the back seat of a waiting automobile, assuming it was a taxi, and demanded "Take me to the corner of Washington and Clay!" Given that Washington and Clay run parallel to each other, that would confuse the hell out of the computer.
In this case, however, the officers driving the vehicle escorted their new passenger to the lockup so he could sleep it off.
If I'm ultimately responsible for the vehicle, I'll stay in control of the vehicle. Because if there's a 10 second lag between when the computer throws up its hands and says "I have no idea" and when the user is actually aware enough and in control, that is the window where Really Bad Things will happen.
Have a look at how collision avoidance systems that are on the streets nowadays currently work:
- the car will sound an alarm signalling probable impending collision and asking the user to intervene.
- the car will also autonomously start to slow down and eventually brake and stop never the less.
The system is designed in such a way that, although human override is possible, the car will also try to autonomously to follow the best course of actions, unless overridden. You could take the control and do something, or you could also let the car follow its normal program (in traffic jams typically).
Same should be applied to fully autonomous cars one day:
in case of "I have no idea" situation, the user should be able to take over control, but lacking any intervention, the car should also react in a sane way ("I have no idea what to do, and instead I'm gona park on the side of the road and wait safely there until further instructions").
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
This is a non-issue because no one actually expected to make a car like Google's prototypes. Google isn't even trying to get into the automobile business with this technology. That's not what they do. They sell tech and software services. They're just trying to show their prospective customers what their tech is truly capable of. The real automakers like GM, Ford, Toyota, etc will use Google's tech in the cars they produce and I don't think anyone ever expected them to make a car without a steering wheel. I'm sure they'll eventually include "automated driving assist" as an option in some of their vehicles but they're still going to sell cars drivable by humans. Don't get me wrong, it did need to be clarified in the laws...but no one was actually planning to build a real car without pedals and a steering wheel.
I want a human to be able to take control of whatever automated device acting as my conveyance. Train, plane, automobile-- it doesn't matter.
If I'm liable for the machinery and the lives carried by the machinery, I want to be able to determine the maximum speed, the maximum rate of acceleration (when not in an emergency), the route, and be able to take full control as necessary. It's not that I distrust computers... it's the squishy meat bags affecting and affected by the computers I don't trust. Humans program the computers and engineer the roads. Sh*t happens and not all sh*t is planned for. There's road kill, potential road kill, flat things on the road that can fly up into the under-carriage when run over-- there are plenty of reasons to be able to control an otherwise-autonomous automobile.
I can predict the future. Due to pressure from Law Enforcement, the new cars
will not have steering wheels and brakes. On command from a Police Officer,
the car will lock the passengers inside, and drive itself to the nearest police
station.
I wonder if you could legally take your driver's license test in a self-driving vehicle. You'd still have to have your hands on the wheel, and check your mirrors before (the car) changing lanes, but I don't know if there are any rules that would actually prohibit your not being in control of the vehicle.
Mind you, they'll have to teach the thing to parallel park...
That is very short term thinking. 15 years after driverless cars are released your going to have a whole generation of people who never learned to drive a car. People who get the license and then forget everything because they haven't touched the wheel in 10 years. There is not going to be anyone quilified to drive a car. You will just have millions of amatuers with a skill level of a 16 year old.
Humans will be maintaining these vehicles and even if they are built to MilSpec they will have power loss and uncommanded control inputs.
Are so many of you so thoroughly indoctrinated by your government into the idea that trading your freedoms, one by one, in exchange for the false promise of 'safety' is a good idea? Because that's what I see too many of you doing on the subject of 'autonomous/self driving cars' that would have NO manual controls, and consequently NONE of you would have the training or skills to drive anyway! You are giving up yet another freedom if you believe this is a good idea.
Are YOU using the TOOL, or is the TOOL using YOU? Think about it!
... also mandates that all driverless cars must be fuelled by petroleum products.
When they've created a Self Driving Car with an automated Wet Bar instead of a Dashboard
If they're insistent there's a way for an occupant to take "immediate physical control", why do they allow current cars on the road?
I'm not sure about steering, but certainly for acceleration and braking there's no way for drivers to take physical control of a modern automobile. Anything we do with those pedals on the floor sends a signal to a computer. The computer then decides what actions to take--open the throttle or apply the brakes.
There's a person initiating those functions, but the person does not have physical control.
I like the idea of a self-driving car, but I still don't understand how the self-driving car finds a parking space, or gets eased into place in the garage for maintenance. How does it find it's way around an unexpected hazard, like a downed limb, or washed-out area of the road? How does the self-driving car know that the road is flooded or otherwise undriveable? How does it know that the power is out at an intersection that normally has traffic lights?
Proverbs 21:19
If a passenger in a diverless car needs to get the car to do some non-programmed maneuvers, they should hit a button, and let the a trained driver back at HQ take control of the car and do the maneuvers.
I am all for a self-driving car. I really hope my current car lasts until we can purchase decent self-driving cars, and I will be happy to purchase one for a price that I wouldn't really consider buying anything else short of houses.
But I want to be able to take entirely manual control back at any time, using a completely non-electronic, fully-mechanical override. I would hope to never have to actually *do* that, but I wouldn't trust any car that didn't have that *available*. Even if I completely trust the pseudo-AI to not bug out and do something crazy, which I wouldn't immediately but certainly could after using it for a while and not having any issues with it... I would *not* trust that some hacker couldn't find some way to hack in somehow and mess with things. Thus, I would want to trust in my ability to perform a manual override and completely temporarily shunt the computer systems off to a place where they couldn't control the car at all.
I am entirely willing to accept the minor inconvenience of hardware UI elements (i.e. steering wheel, pedals, etc.) getting in the way of aesthetics, and the slight additional cost of manufacturing said UI elements. And I'm *glad* they're pushing for that to be law, or else nobody would actually do it, even as an option - we've seen a number of examples of things I'd love to be able to pay for (phones with physical keyboards, 16:10 laptop screens), but nobody makes them because they're slightly more expensive to make, and companies feel the drive to make everything as cheap as possible and not even give the option. :(
That Google thinks their self driving cars are ready for the open road isn't the issue. The issue is that they think they are ready to go straight from traditional cars to cars with no ability for the human passenger to take control if the new, unproven technology fails. That, by itself, convinces me that Google's judgment is flawed, and cannot be trusted. Were I making this decision, I wouldn't let Google's cars on public roads at all until they show some evidence that they understand why this is a bad idea.
It will have to be more of a kills switch than a break pedal...I would think that you would want a braking system installed in the car that if the CPU signals were ever interrupted, it would automatically engage an manual break system. The 'break pedal' would then be a kill switch that turned off the computer controls, and activated hazard lights. This would give you an out in case the car CPU failed for any reason, or the passengers panic.
HA! I just wasted some of your bandwidth with a frivolous sig!
what about utility trucks / bucket trucks?
They need to go off road (off road can just be on to the grass / sidewalk next to an road) may need to sit in the middle of a road, be placed the wrong way on a road, etc. And it may be hard to get a auto drive car to do that.
Also based on way some utility's (manly cable) use contractors and subcontractors a special license sounds like a good idea but in real life may not work to well. And they may need to make so they can't use contractors and subcontractors do that work or they may have to make so the utility / cable co has to buy the special licenses for the contractors and subcontractors.
From a legal standpoint,
If Google makes a car and it gets in an accident, who is responsible
Google, or the operator of the car?
The lawyer's answer is of course both, and especially whoever has deep pockets.
For driver-less cars to become real, we need a better, clearer set of rules.
For a long time, it needs to be operator is responsible, so he needs manual override.
From a safety standpoint,
It may be that a fully autonomous car on a freeway with only other autonomous, communicating cars is way safer than a good, human driver.
For unexpected things like a mechanical problem, an ice patch or deer crossing, this is not completely clear.
Barring reserved toll roads, a sterile roadway is not going to be the case for a long while.
A fully autonomous car on a freeway with unpredictable, non-communicating human drivers seems at best an experiment in what new features the car needs.
The operator really needs to be on top of his game when other cars are near.
Operators able to handle hours of boredom followed by seconds of panic are rare.
No manual override for now is folly.
There are still a lot of situations where the automatic controls would not know what to do without a second thought. I live in a part of the country where it snows. Evidently when road markings are covered by snow, the automatic controls don't know what to do. I am expected to be at work on days when it snows, and I can drive on a snowy road without much of a problem -- I just need to take into account the variables as they present themselves. Another situation that pops up occasionally is that I need to park the car in a field or other unpaved surface -- there are no lines and no dedicated spaces, and I'm often directed where to park by a human. I can't imagine the self-driving car knows how to do that yet.
Not to undermine in any way the fantastic work that's been done by Google and other people working in this field. For many conditions that are well-understood (e.g. long distance freeway driving on a clear day), it seems like a lot of this technology is nearly ready to come together in existing automobiles, where you put the GPS, lane departure and adaptive cruise control together and pretty much drive the car. I even saw a video of some guy letting his Acura pretty drive itself with these technologies enabled. He was a total flippin idiot for actually getting out of the driver's seat, but the car did everything right. Video here: https://www.youtube.com/watch?...
Technologies in the wrong hands can be made to be wreckless and dangerous. What assurances do corporations today give the public that their technologies won't end up in the wrong hands? I see about 5 people a day texting while they are driving. I doin't even get out much. Corporations don't base their business models on how irresponsible someone will be using the stuff that's made for profit. Google hands spy satellites to the entire world to use as they will, disregarding the fact that there are severe lines drawn by warring nations. It hands spy glasses to the public with no concern for people and their right to privacy even in public places. I can't wait until corporations like google and facebook are no longer alive and well on planet earth. The day is coming. I might not be here, but I rest assured that their days are numbered.
I'm surprised they even tried to go without manual controls. There will ALWAYS be situations where a human may have to take over.
1.) Safety override. 2.) Driving in an open field or on the beach without a precise destination. 3.) Parking garage, car wash, mechanics bay or other situation where it MAY be necessary to maneuver the car in a way that would be impractical to tell a computer what to do. 4.) Weather conditions could prevent automated operation. 5.) The ability to violate laws when an emergency situation demands it.
The driver must always be ultimately responsible for what happens while they are behind the wheel, a person cannot be accountable without control.
I can't believe no one here has figured this out. Computers don't and can't know everything about our ever-changing world.
If your car does not have a user-controlled steering mechanism, how can you ever drive off the official road (in a good way)? If I have a private road do I have to survey it and register it with the government and then wait months for it to be properly entered into their database before I can drive down it? If I have a farm would I no longer be allowed to drive out into one of my fields? There are millions of unmarked/unregistered/unnumbered driveways off secondary roads. Can I no longer use any of these? Is my self-driving car going to be smart enough to know when it is okay to park in a garage, and if it is a multi car garage which side to use?
Databases *always* contain out-of-date or wrong information. As recently as 2 years ago my TomTom insisted that I should drive the wrong way down a one-way road. Also just a few years ago Google had the old (now wrong) information for some roads that had been reconfigured and moved 10 years earlier.
I can easily imagine other cases where I can see and understand a hazardous situation before the car can notice it. If I wait for the car to notice, that could place all occupants of the car in unnecessary danger. If a deer is running toward the road and I am in control then I can slow down and let the deer run in front of me. Is a self-driving car going to always handle a situation like this better than a competent human?
Finally, say the driving AI computer fails. Would that mean I could not use my car for potentially days or weeks until I shelled out the big bucks to have it replaced? Right now I can have dozens of non-essential parts fail and I can continue to use my car until it is convenient for me to fix the problems. The engine, drive train, and wheels are essential parts. An AI system is not.
...For maximum safety, a manned car should drive ahead of the self-driving car at a distance of no more than 200 meters, with an illuminated banner that reads "Caution! Self driving car follows behind."
Time to draw a line in the sand - in 10 years no manual cars will be allowed - end of discussion. Commit to it and watch whole industries be created and. Yes whole industries will fail (red light cameras, traffic lights, traffic cops, etc.) but those are just bumps on the road that need to be end of lifed anyway.
Those who sneer at the idea that steering wheels and brakes need to remain on self-steering cars should watch some of the airplane accident videos on Youtube. You'll see case after case where the pilots bring in a plane so badly damaged, the plane's computer's are clueless.
Perhaps the best illustration is a state-of-the-art Qantas Airbus A-380 where an exploding engine took out so many systems and instruments, all the aircraft's sophisticated computer system could do was spit out dozens of error message. In that situation, the pilots cobbled together enough control to bring it home, badly damaged and heavily overloaded.
https://www.youtube.com/watch?v=s29OI5VeW30
It's a great documentary. I'd trust an experienced pilot over any computer ever made. That pilot wants to go home. The computer doesn't care. For it, existence is just bits and bytes.
I must remark about this silly posting: "don't want to let untrained people having the ability to make things worse."
Untrained? This is a car with a out-of-control computer that can kill in a few seconds. It is not a complex airliner in a crowded airspace. And if you want to take control of that car, you need to take over steering, braking and acceleration. One or two isn't enough. It's stupid to think otherwise. Steering will keep you from smashing into a concrete wall. Braking will end a mad drive. Acceleration (and the others) will get you to a hospital despite an unwilling computer. People always need complete control.
One of the divisions in humanity is between those who think most people are competent to run their lives (generally conservatives and libertarians) and those who think they lack the proper "training" and need to be run by their betters (liberals). This is a good test for that.
Give me a complex device like a ship, aircraft or vehicle to manage and two questions will stay in my mind:
1. How to deal with failures? "I A fails, what can I use?"
2. How to override or force to shut down anything the device might attempt that's dangerous? Think of out of control accelerators on some recent cars.
In short, I run it. It does not run me. For illustrations of the opposite, think of HAL in 2001 or Skynet in The Terminator.
And they will be taxed twice as much.
I'll stick to my opinion that this self driving thing will mostly be a fad. For ~10-20% of the population who really can't drive (elderly, vision impaired, etc) it makes some good sense, then maybe another 10% who have long commutes that rise to the level of being a burden. The other 70-800% simply won't see it as a worthwhile capability in the next couple decades.
My suspicion is that the real utility will be in automating 18 wheeler traffic and with improving public transportation. General utility just does not compute for this curmudgeon.
Besides, I can't wait to see what the EULA looks like for one of these things.
I agree with the numerous posts that it's a farce to think a human driver could possibly be fully at the ready to take control in the event of an emergency. But I still don't see human control as useless. If a complex situation comes up, say involving construction and cones and unusual lane changes following the cones or whatever, the self-driving car could simply report that it doesn't know how to handle this situation and come safely to a stop, and hand control to the human. I think that's where the value is in having human control available as an option.
If we want airline levels of safety then cars will be drastically different. The autonomous vehicle cheerleaders CHOOSE not to mention this, I suspect because many of them hate human operators and have an exaggerated faith in technology rarely shared by people with actual experience working on high-tech vehicles.
I'm an aircraft mechanic but most people are not, and most people would shit a cactus if they had to pay to maintain their car to aircraft instpection and maintenance standards.
Autonomous cars and trucks will have to have a comprehensive sensor suite to warn of pending failure and in some cases disable the vehicle so it can be safely towed off for repair. A conventional auto can limp to safety using manual steering and brakes. Those systems also operate even with total loss of electrical power. They are highly reliable. Since owners who are not technicians cannot competently assess system conditions, the automated systems must take away their choice and that means some control of "re-arming" the vehicles for legitimate test and diagnosis. Built In Test systems aren't perfect even on jet fighters.
New operation and maintenance regimes must be devised that ensure sufficient reliability to take away manual controls AND also have a degraded but useful level of failsafe operation if power is lost. For example, braking systems should default to "applied" if power is lost rather than coasting to a crash. Commercial truck air brakes already do this if pressure bleeds down. Springs apply the brakes if system pressure drops to a level which won't hold them off the drums/rotors.
They will also need to be EMP hard. Not because of some apocalyptic nuclear scenario which isn't going to happen, but because otherwise they will be easy kills with primitive HERF equipment, and even more so where metal bodies give way to composites and plastics which offer no grounded shielding.
HERF kill a conventional auto and they can be steered and braked to a stop the same way as when they throw a rod or seize an engine because there is actual mechanical and hydraulic connection between the meatsack and the machine. HERF a major highway during rush hour and there had better be mechanical defaults to cover degraded and disabled electrical systems.
One major reason aircraft don't drop out of the sky very often is component time change requirements. Healthy components are replaced BEFORE they fail. It's very expensive. For that to happen on autonomous vehicles it must be mandated by law, and as component life expectancy changes or recall-worthy items are found the discretion to replace "safety of flight" items cannot be left to Joe Sixpack. Of course owners will rage at what they consider legislated theft.
There is no option not to deal with the above concerns and it will be entertaining to see how responses are thrashed out.
I can see the early days there being "enhanced cruise control" type thing that people can be licensed for that can be used on freeways only in the near future. I would basically be cruise control that can brake and accelerate and then maybe steering capabilities included, etc.
I don't think it would be possible to have some mix of fully autonomous cars with regular ones on the road at any time.
Anyone who thinks self driving cars are likely to be capable of driving on open roads in all circumstances by themselves in the forseeable future is living in cloud cuckoo land. There MUST be a conscious, unimpaired human being able to take over when the need arises because the need will arise.
Convince the AI to pull onto your yard to wash it :)
I'm an aircraft mechanic but most people are not, and most people would shit a cactus if they had to pay to maintain their car to aircraft instpection and maintenance standards.
Not necessarily. There would be economies of scale. There are also a lot of practices in aircraft maintenance that could be streamlined, like having central records so that every mechanic isn't responsible for basically double-checking everything every mechanic beforehand has done. It shouldn't be any more expensive than just dealer-maintaining your car over its lifetime, which is WAY cheaper than a typical aircraft maintenance regimen.
Autonomous cars and trucks will have to have a comprehensive sensor suite to warn of pending failure and in some cases disable the vehicle so it can be safely towed off for repair. A conventional auto can limp to safety using manual steering and brakes. Those systems also operate even with total loss of electrical power. They are highly reliable.
I'm sure there will need to be a differing set of standards for redundancy/etc for reliable autonomous vehicles, but I'm not convinced that it would be over-the-top. How often do cars have major electrical failures? I imagine that it happens less often than on your typical 40 year old Cessna, even if the results are worse. You could also have dual electrical systems and such, especially for critical components. The car really just needs to be able to coast to safety - this isn't a plane where no radios in IFR is a potentially deadly situation.
Since owners who are not technicians cannot competently assess system conditions, the automated systems must take away their choice and that means some control of "re-arming" the vehicles for legitimate test and diagnosis.
I do agree with this. However, in many cases the cars could also drive themselves in for PM well before they are crippled, which would be a major convenience. People would be far more likely to rent and share cars as well when they can be at your door in 2 minutes.
They will also need to be EMP hard. Not because of some apocalyptic nuclear scenario which isn't going to happen, but because otherwise they will be easy kills with primitive HERF equipment, and even more so where metal bodies give way to composites and plastics which offer no grounded shielding.
I don't buy this. It is a bit like saying that conventional cars need to be hardened against 50 cal sniper rifles, since one shot to the driver could cause a pile-up. Shooting an HERF at an autonomous vehicle is homicide. So is firing on an airliner during its takeoff roll - I'm sure the same sniper rifle would cause a lot of havoc there. The solutions to these problems don't involve systems engineering, though I'm all for reasonable levels of robustness.
Healthy components are replaced BEFORE they fail. It's very expensive.
Actually, on airliners many components are flown until failure when they aren't critical for safety and there is sufficient redundancy. This is part of what makes them more reliable than general aviation - they aren't constantly doing things like servicing all the magnetos (we had a fatal helicopter crash nearby due to improper maintenance and resulting dual engine failure a number of years back).
I do get what you're saying though. If we really want vehicle safety to rise to the level of flight safety, then you can't have people doing their own oil changes without the equivalent of an A&P.
Let's face it, this is how it will go.
Firstly the cars will have dual controls. Eventually they will have no controls. It is inevitable.
It is a sad thing to face because I actually LOVE driving. I enjoy it. I don't care whether driving my own car is slightly more dangerous than having google drive my car for me. The bottom line is that I REALLY ENJOY DRIVING AND I WANT TO CONTINUE. I purposefully drive a manual car. Why? because I get enjoyment from driving and I get the most enjoyment from driving a manual car!
Not to mention that I want to be able to be in control of the vehicle. I like it. Let's be honest, we love driving and google is trying to take it away from us.
It makes me sad to think that my grand children might never experience the pleasure of driving a manual car, that they might not experience one of the modern rights of "passage to man hood" of obtaining a driver's licence. It really does sadden me.
You're thinking about flying drones, not driverless cars. Depending on the drone it pretty much varies between the remote operator actually flying it all the way down to simply programming a flight path that the drone then uses to take off, fly, and land without further intervention. Most military drones do have plenty of intervention, but again, that can range from taking over and 'flying' to simply adjusting waypoints.
I don't read AC A human right
My understanding - perhaps reading on /. - about human caused crashes (Asiana, Air France) was that the plane did so much for the human, that the human was more out of touch with actually *flying.* A cautionary tale if we're thinking of handing over control of cars. Humans will get out of touch. Maybe the right place initially for driverless are paid express lanes with limited entrances and exits, and perhaps solely one lane, making a lot of the scenarios discussed by other posters less likely initially.
In a decade or so, cars will likely be able to drive with no people on board, or even transport children with no adult in the car.
The driverless car is not a chauffer or an nanny.
It knows only that it has arrived at its pre-programed destination. It doesn't know whether that destination was entered correctly. It doesn't know whether that destination is currently a safe place to drop off your kids.
The state of AI is such that we are nowhere close to driverless cars. Lets cross this bridge when we get to it.
Let's first get rid of the clutch from modern vehicles, then we can consider removing steering wheels / accelerator from these ones...
So what exactly is the difference between a Google car without a steering wheel and a drive by wire system used in some cars today ?
I think it is sort of like with a lift. Initially there was operators and all kinds of safety requirements but today we are so used to using them that most of us don't even know how to open the doors in an emergency.
Wouldn't hurt to have some way to check blood alcohol level before allowing a human to take over.
blindly antisocialist = antisocial
Just make sure that the main power cord can be disconnected by an occupant at any time: well designed software should always know how to fail gracefully and in this case a loss of main power should see a series of safe-shutdown routines kick in, appropriate to the context
Cars will always have a human driver. The manufacturers don't want to get sued for every wreck...
Remember that it's against the law to leave children alone in a home under a certain age, and alone at night at a higher age. I can't imagine it being okay to have children alone in a self-driving moving vehicle, with their door locked to prevent them from exiting in an emergency, and no adult to help them.
People need to take control when they think the Google driving is too slow or whatever.