Wanna run someone over in your next Ford? No dice, it won't let you Ford has been flaunting the latest gizmos it plans to add to its cars, including a self-parking mode designed to squeeze motors into smaller parking spaces. At a testing facility in Germany the company has been showing off the latest gizmos that have been built into one of its Focus cars. The vehicle is equipped with radar, …
My car has this, and does not react to plastic bags. Mind you, I have not noticed it do anything, ever. That may be because I've already taken the action needed or because it doesn't work. It seems churlish to test it by getting my wife to jump in front of the car and see if I run her over or not.
Don't forget these things are only set to do stuff at relatively low speeds, reducing the possibility of the rear-ending.
A rear-end shunt is always the problem of the tailing car. The driver should leave an adequate gap. Of course, no consolation to you if you are the chap up front.
"Of course, no consolation to you if you are the chap up front."
Or, in my case, the chap behind - someone recently rolled back into mine on a hill, when a queue of traffic began to move again.
You might think I didn't leave an adequate gap between my car and the one in front, but I did: She actually managed to roll back four sodding times!
Luckily for her no-claims discount, there was only minor cosmetic damage that I'm not going to worry about.
I expect most people who passed us on that hill probably assumed that I had driven into the back of her car.
AC about the 'hill roll back' comment....
I remember I traded my old automatic car in for a shiny new car with a 6 speed gearbox.
Being a 6 speed, reverse wasn't traditionally below 5, but you pushed down and selected 1st.
You can see where this is going...
Hill start, new car, panic, lean on gearstick, revs, take off and it starts going backwards!!
I'm very lucky that the car behind left plenty of room!!!
"Your car automatically avoids that plastic bag that just blew in from your peripheral vision and steers into a concrete bollard instead...."
That's awesome, how people on this thread spend 2.5 seconds thinking of some idiot scenario where the system might fail (plastic bag, swerving into oncoming traffic to avoid hitting a deer, etc.) and then assume that Ford engineers all have the IQs of small dogs and haven't though of any of this and designed the system accordingly.
> Your car automatically avoids that plastic bag
Unless the plastic bag sufficiently matches the IR signature and shape of a person (or animal, depending on the system), then the car will not worry about it.
> and steers into a concrete bollard instead....
The system only controls braking, not steering.
Any other insightful contributions from you today, Sir?
This does make me wonder if it would result in frying pan into the fire incidents, like taking avoiding action by going into oncoming traffic (or in some cities, a tram)... Just so long as it doesn't encourage people to let their guard down by not paying as close attention to the road around them, thinking the car will save them.
That's my first thought as well. I just find it a bad idea to take control away from the person who's life may depend on having that control. When I was working on motorized wheelchairs there was a short but decisive debate about giving whether to enforce a stop or reduce performance at a critical motor temperature until the motors cooled back off to prevent the motor burning out. In short we decided in under a minute to warn the user about overtemping the motors but ultimately to let the user burn out the motor if they so chose. The reason was we could never know if the user was actually trying to escape a burning building where it may be impossible to cool the motors.
> That's my first thought as well
Ito, given your alleged experience on wheelchair design, where you have been exposed to analogous problems, what makes you think that the engineers who designed this system (the name of the designer / manufacturer escapes me right now--no, it's not Ford) are any less astute than your team?
For info, I have a similar system and it never "takes control away". I am perfectly capable of ramming into the car in front by taking positive action like stepping on the gas. It's only if I do nothing that it will try to slow down/stop, and that only at a distance that is so uncomfortably close that any non-incapacitated driver would have slammed the brakes by reflex well before the thing intervenes. It will also, I'm told, aim to stop as close as possible to the obstacle in front, so as to give following drivers plenty of braking room.
Other teams also do think things through sometimes. I respectfully suggest you go and try one of these systems before formulating opinions on the basis of non-information. After all, you wouldn't like people talking bollocks about your own design either, would you?
@AC 20:49
"what makes you think that the engineers who designed this system are any less astute than your team?"
I don't know what you read but I did not state any such thing. I merely agreed with the original poster that the frying pan into the fire scenario was my first though as well. I never said it was my only thought but don't let that stop you from claiming that I'm "talking bollocks" when it's evident I'm not. I stated that I think it's a bad idea to take control away from the operator and merely stated my experiences. If it can be manually overridden then it doesn't fully take control, does it? It may be that a simple flick of the wrist will release the system back to the driver but that isn't clear since overriding the system isn't shown in the linked video. Granted, it's partially implied since the driver has his hands very near the wheel so as to catch it if it goes wonky but it isn't known if that is just a technology demonstrator in debug mode or something much closer to the final product. I also understand engineering, especially the research and development variety, and it comes with lawyers, lots of lawyers who not only obscure the language to shift as many patents as possible but also ones who understand that anyone hurt by the latest miracle will be launching massive sueballs at the company.
Since you mentioned non-incapacitated drivers, what do you suppose it does with one who is incapacitated and the car decides to swerve instead of stop as shown in the video? Does it then proceed to stop or wrongly release the car back to a person who is incapable of controlling it? I don't know. I also don't know if the engineering team thought of it because they didn't mention they had so I must assume they have not and would prefer they think about it now that it's been brought up. As a result this is a great technology demo but we don't know if it will make it past the lawyers to become a real product or even if the U.S. government will go ahead a make it mandatory like they did with tire pressure warning lights, traction control and soon rear view cameras for reversing.
"After all, you wouldn't like people talking bollocks about your own design either, would you?"
For your edification, I have these things called design reviews. The purpose of these is to do our best to determine if we have thought of all the details such as was the case with the wheelchair. In short, no I don't mind at all if people want to point out potential deficiencies in my designs, in fact I prefer to get such feedback as early and as often as possible. I'm not anywhere near that arrogant, as you so obviously think.
Unlike the driver on occasion, systems like this scan the whole road ahead (in this case for up to 200m) so the evasive action it takes is not just to dodge the immediate threat (like a reflex action that a human driver might take) but to avoid the immediate threat by moving into a space that is not threatened. So surely it is likely to be safer than a person, not just plow you into oncoming traffic to avoid a cat.......
But 200m down the road I can still see an idiot hurtling along at 90Mph in a 60,
it can't, so a dear runs out in front of me, normally I would break, instead this thing steers me around it into the path of the mad driver it could not see...
I have forward alert on my car and it is great, because it does warn me of potential collisions and I react to break, there are times when you need to check blind spots, and that is when your eyes are not forward, and that is when forward alert is great.
But I would not trust my car to change lanes on me...
I've tried auto parking on different cars, it is shit..... I can park in spaces MUCH tighter and more accurately than it can...
Even if your 90mph idiot is there, as soon as the on-board system detects you are in line to hit them (at it's 200m range) it avoids him too - if you are doing 60mph and the idiot is still doing 90mph your closing speed still gives the system around 3 seconds to get you out of his way, at 60 mph you will have travelled 80 metres in that time - more than enough to get you out of his way too.
The point being that the system (like a Terminator) sees everything all the time, not like a human that sees pretty much just what you are concentrating on and little else - it's not perfect but chances are it's better than you (or me).
The issue is eloquently expressed when you say 'I would not trust my car to change lanes on me' - it's the trust of the device that needs educating. People are quite happy to trust a car to use ABS appropriately or brake force distribution or stability control - all these things can significantly alter your vector without your control, so why is the steering special ?
Granted, there might be very, very, very unusual circumstances where the system gets it wrong, but killing one person by mistake using a system that saves 999 others by its operation is surely worhthwhile odds.
Errrm, hang on. So your car has moved you into the oncoming lane to avoid a car or person in your lane, so your in the wrong lane now, then it detects that 90mph idiot on his way, so where does it go then? It can't necessarily go back into the correct lane as there may be traffic ahead in that lane or you may not have passed the obstacle you originally avoided. So what does it do? Take you onto the pavement to kill a pedestrian or let you hit the 90mph car head on? Personally id rather have stayed in my own lane and attempted to stop in time. The system just can not account for what it is going to get you into by taking this action and who is liable, Ford or you?
Nothing about swerving into oncoming traffic or off into the ditch. It just applies the brakes.
Yeah, I thought the same, but I followed the supplied links and found that it will in fact take over both steering and braking as needed. I still think the system Nothing about swerving into oncoming traffic or off into the ditch. It just applies the brakes.more likely to react better than most drivers to surprising situation. To get back to the point concerning a deer wandering into the road, so very many people swerve when dealing with one of these beasties, and end up hitting a tree instead. Since this is one of the most common situations in which something fairly large wanders into traffic, I would hope that Ford tested for exactly this.
Unless you're the one.
I don't think that the "needs of the many" argument holds much water in court either. If you run over someone you're still guilty of manslaughter. Saying that the system didn't do its job is not going to get you out of jail.
@Pascal Monett - so your contention is that no new safety system should be introduced unless it can be shown to be 100% risk free even in a highly unusual scenario ? Yes, the one dead person is still dead but by that logic we should go back to 1950's style cars with no safety features at all because they can all lead in some way to a death (seatbelts can entrap you in a burning car, airbags can damage your face or hearing etc.) - The "needs of the many" argument holds plenty of water in court as long as the manufacturer can demonstrate that the system is made as safe as it could be and is overall beneficial.
It would appear, also, that many commentards are clearly ninja level drivers, on account of them being much more aware and fast than a computerised system despite the fact that in the crash scenario under discussion their vaunted skill has already got them into a situation where they are having to choose between hitting a car in front of them and hitting one coming the other way and have failed to react to that developing situation for so long that the computer has needed to intervene.
In such a situation, the driver "in control" of the car has already failed to act properly. They are not at the point of making a sensible decision about choosing the lesser of 2 evils by bumping the car in front instead of pulling into the wrong lane, they are about to have a hideous crash. The computer is perfectly able to judge which of 2 definitely happening crashes will be the less serious and opting appropriately. It is more likely to make the correct decision than the driver who, as a human, is more likely to choose to avoid the immediate (but less serious) crash because it's right there in front of them and their reflex will have made the dodge before their poor old brain has even realised it's happening.
So the question seems more to be about whether you would prefer to kill 10* people but retain that human touch or kill one but have it done by a machine. I suspect the 9* saved by the machines might have an opinion on that.
*Actual numbers made up and not relevant to the point - if the computerised system is even 1 in 1000000 safer than the human option, then it still wins.
The issue is eloquently expressed when you say 'I would not trust my car to change lanes on me'... all these things can significantly alter your vector without your control, so why is the steering special ?
Steering is special because it can typically "alter your vector" at a much higher rate than differential braking pressure can since the former can do it with nearly zero tire scrub while the latter must add considerable drag to change the direction of the vehicle. Add to that the traction control systems also expect the driver to, well, drive. Also the function of traction control systems are to compensate for simple mistakes by the driver, such as stepping on the brake or accelerator too hard, in order to help the driver maintain control not take control away. In the end, differences in tire pressure can change cause a car to drift too but I'm not about to sweat the course correction required by having an extra 2 psi in the left side tires.
The other thing this system ignores is the fact that there exists an enormous unknown and that is the reaction of the other party or 'target' if you will. Should it assume that since it has changed lanes only to find an oncoming idiot going 90 that the idiot will remain on course or that it will steer into the target it has avoided? Will the idiot have the same system and both systems respond by juking into any open space only to find that space occupied by its equal and they juke and jive back and forth until they meet in the middle? Granted it should be applying the brake to reduce the impact but a reasonably skilled driver might also pick the lesser of two evils and switch back into their original lane and I doubt a computer ever would.
Finally, I don't really mind the braking when something darts in front of you so much but assuming the system has previously warned the operator of a potential collision or hazard, which it could easily do in the scenario shown in the linked video, why doesn't it then proceed to make a safe stop? It has decided the operator is deficient for not reacting to the warning so why shouldn't it come to a complete stop as safely and swiftly as possible? It has no way to know if the operator is awake or suffering a medical emergency such as a stroke or heart attack so why would it ever give control back without some sort of confirmation from the operator that they are able to resume control?
Sure, they might save people's lives some of the time, but they have two major downsides:
(1) They confuse responsibility. The driver should always be responsible for all actions and outcomes of all acts in the car. Now were confusing that picture by stepping in with technology. Instead of checking there is no kid crawling around on the driveway, the driver takes a chance that the technology does not work. It won't always work.
(2) Even though a device might save 999 for every 1 it kills, that 1 will get the car makers and the engineers in court and get their pants sued off. The 999 that were saved are non-events. That 1 that dies is a newsworthy event that the media and lawyers will feast on.
A glaring example of this is Therac25 which, over a period of 2 years or so, nuked 6 people, killing three and earned itself a place in the cock-up hall of fame/shame. This cost the makers millions in legal fees and damages and no doubt tainted the engineers involved for years. Nobody remembers the fact that during the same two year period thousands of people were treated and their lives were saved.
"I've tried auto parking on different cars, it is shit..... I can park in spaces MUCH tighter and more accurately than it can..."
So can I, but it's pretty clear that there's a large number of people who can't park for shit.
Self parking is NOT going to stop people double parking, parking on double yellows, on corners, footpaths or on the wrong side of the road (yes, it _is_ illegal in the UK, unless park lights are left on) but it'll go a long way towards sorting out the wankers who leave 3 metre gaps front and rear.
No disagreement with the principle. That said I have found that obeying traffic laws as a cyclist does bugger all to improve how other road users behave. I still get tail gated when following a slow car, I still get cut up by folks making left turns and treating the gap I am maintaining as a handy place to put their car while they brake as hard as possible to avoid running into the car in front.
In my experience someone who rides a bike in an inconsiderate, dangerous and illegal manner will also drive cars in exactly the same way. The fact is some folks really should not be on the roads full stop and given the choice I'd rather they rode bikes than drove cars because they can do a lot less damage to those around them with a bike.
> Running over? Abso-fucking-lutely. "Share the road" means cyclists need to obey traffic laws, too.
You know? I drive "premium" cars, the kind that most people assume we think that road laws do not apply to us. As far as I'm concerned, though, the cyclist is making progress by means of his own physical effort, exposed to the elements and, more worryingly, to motorists and other hazards. Why shouldn't I show some deference to them, and be tolerant of any minor transgressions that may occur in the interest of safety or efficiency?
"...running a load of said motorcyclists off earlier up the road."
Perfectly untrue. It's unlikely that Mr Lien will even get a ticket.
The bikers were out looking for trouble. They found it. I think that they might have forgotten that they were on motorcycles and were 'taking on' a Land Rover. Idiots.
"Perfectly untrue. It's unlikely that Mr Lien will even get a ticket. The bikers were out looking for trouble. They found it. I think that they might have forgotten that they were on motorcycles and were 'taking on' a Land Rover. Idiots."
So far, apparently however public opinion is begining to question the events, statements made by the police and other factors, and now Ray Kelly is stating its possible Lien will face prosecution for his part in it too. Apparently that group of mindless thugs also included undercover police and some off duty uniforms too.
Meanwhile, I read the comments left by most "normal" people about Mieses and how he deserved it (even though it was Cruz who triggered the incident) who was crippled while parked up and attending to someone lying injured on the floor and trying to calm a situation down and see people baying at how they would run everyone else over in the same situation, pull out guns and plug a few people full of lead etc.
And for my shame I actually drive a Land Rover so risk being tarred with the same.
Posting anon because I am completely aware that my reasoning offends "normal" pre programmed people from clearly stated non shouty discussions in my office at the coffee machines. Such is the price of media hysteria. And I have to keep my role while saving up for some oasis as far away from "civilized" society as possible.
> (or practicing sliding/fishtailing on the snow/ice in un-occupied parking lots).
In other words, being a wanker on someone else's property? I hope you'll be kind enough to pay for any damage caused, e.g., next time you run into a lamppost.
There are facilities specifically for practising this kind of stuff, with experienced instructors at hand for a very affordable price.
"Self-parking cars are nothing new. Volkswagen showed off a self-parking car in 1992, Toyota has its Intelligent Parking Assist System (IPAS) that comes with some Lexus and Prius models, and Audi and Volvo have also got similar systems. Ford hasn’t given a data for the release of its try in the area, but has promised some news by Christmas."
I think Ford's version is already out. My mother's 2013 Ford Explorer has Active Park Assist and it works very well. I had to give it a try when I was visiting the family last month.
in those high speed Hollywood action movies, or Zombie apocalypse stories unless the sensors can tell living from living-dead and will let you plough them down.
On serious note - will the parking sensors leave enough gap between parked cars so that you don't box others in?
> On serious note - will the parking sensors leave enough gap between parked cars so that you don't box others in?
No. The driver is always in control of the throttle and therefore it is him who decides how much space to leave. Park assist systems will just give out the usual proximity indication from parking sensors (which are a requirement for these systems to work) or rear/front cameras if installed.
Btw, I tried the Ford system on a hired car a few months ago and it was rather poor at identifying kerbs so it would often park with a wheel or two on the pavement. Also, I did not find a way to select parallel park. Can it only do inline?
I have also tried the system available on Volkswagens (another hire). It performs much better at both spot detection and manoeuvring, and can do both parallel and inline parking. Still a waste of time for any skilled driver, mind you, but I understand why it is useful to some people (not to say they're bad drivers once on the move, just bad parkers).
To tell the difference between a person, and say, a badger/pheasant?
The last thing I'd want is a lift off oversteer inducing moment on a muddy backroad because the car tried to steer my around an object it'd be far safer hitting, all things considered - I got a pheasant the other day at 50mph - it literally sprinted into the road a few meters in front of me. it got safely (from my point of view) punted into the next field like a drop kick, the only mark on my car beings it's bowls evacuating as it bounced off the bumper, which is obviously made of sterner stuff that I feared at the time!
Even something like a sheep - I'd rather slow to 15mph and twat it than potentially put the car into a position where it can slide off the road due to aggressive steering on a slippery surface into a ditch - or on some of the roads around here, some nasty ravines....
As long as it's switchable, I'm happy, basically. The number of people who just step into the fecking road in towns these days is staggering considering how batshit mental an act is is to do it. If some kind of MAGIC can stop me from being blamed for someone elses stupidity, and I can switch it off when I'm in the dirtier, unkempt, poorly cambered, muddy backroads, I'm all for it.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
