Ford's parallel PARCing: Motor giant tries to craft new tech just like Xerox Ford is scared of the future. It has to figure out how to survive in a time when manually driving your car to work is as archaic as commuting by horse. The motor company is running research projects to look at how the future might be multi-modal, with car-share schemes, folding electric bikes, and self-driving cars. There is a …
I think the media is far too obsessed with self driving cars something that is generations away. yes we have concepts but even cars new today take decades to cycle through the market.
The problems presented by self driving cars are not insurmountable but we are a long way from a solution. This is not cruise control on a plane or a ship this is fully fledged artificial intelligence using visual processing.
@Terafirma-NZ
Yes - because there are no cars which can do this at the moment.
OK, they can't yet do it in all weathers, but the GCar regime has shown it is well within our capabilities - and there are others out there.
I don't think this will take decades - I'm hoping that my kids will never have to learn to drive (they may well choose to though)
You point out that "we have concepts but even cars new today take decades to cycle through the market". So ? To make it happen someone needs to be working on it now. That someone might as well be Ford since they make cars. And the media might as well report on it. Im not sure that I see your point?
But there are millions of cars already on the road. Autonomous cars are only 5-10 years away, but like the introduction of ABS and cruise control it will take a lot longer for it to trickle down to the cheaper cars. Initially it will be the top of the range that get this tech, but cheaper cars will only get them as older cars are sold as 2nd hand.
There are a couple things that are going on...
First there are some new tech like the LED fiber that can be woven in to fabric... That could be very interesting.
Then there are some brain dead things like trying to map out open parking spaces on the side of the road using the sonic parking assist sensors. Just a couple of nits with this...
First your map data is at most accurate to 1.5 meters. So mapping the data of an open spot could be a bit of an issue. If you're on a road where you have parking and just a single lane of traffic... ok. But if you're on a 4 lane road... think about that for a second.
Then you have the issue that there are conditional parking rules. You need to be able to read the signs which require photograhs and then the software to figure out the shape, and then the OCR of what's written on the sign. This is a bit more problematic. (Loading zones during business hours. Valet parking only... temp parking.)
Then there's the next problem... a Ford drives by... see's an open spot. You're two minutes behind on the same road. But by the time you get there... its gone. Other drivers... So the data has value that is fleeting.
With respect to remote driving cars... Talk about a major problem there. Sure we can put in a sensor to make sure that no one is in the car at the time... but what happens if the remote driver decides to use the car as a battering ram? Causes traffic accidents, etc ...
While some things are technically feasible.. that doesn't mean that they are a good idea.
"
Then there are some brain dead things like trying to map out open parking spaces on the side of the road using the sonic parking assist sensors. Just a couple of nits with this...
"
This struck me as a great idea that is completely realisable today. Most of your objections could be easily overcome by mean of a comprehensive updatable "parking" database, along the same lines as existing satnav databases. My satnav "knows" what the speed limit is for the road I am on despite not being able to read a single road sign. It can also direct me to the closest fuel station or restaurant and warn me about most speed cameras. The data on a satnav that has the proposed "find a park" mode would include parking times, restrictions and costs. I can think of several ways for a system to deduce whether or not you are in the outermost lane. Sure, it could not guarantee that someone else doesn't get to the parking space first, but it would still be far better to have directions to the closest 5 parking spaces than driving around slowly looking for one.
"My satnav "knows" what the speed limit is for the road I am on despite not being able to read a single road sign."
This is because one of the attributes of the road link you are traveling on is the speed of the road, along with the classification of the road link and other data. Its static. Now if you want to talk about capturing the speed limits on roads with variable speed limits ... that's a different story.
Your gas station is a PoI (Point of interest) that is a layer on top of the raw map data. (Well it could be part of the map data depending on who's GPS system you are doing...
Now you have to ask yourself how did your map data get so smart? How did the speed limit of the road link get captured? ;-)
Again I'm posting anon for various obvious reasons.... are you sure you want to argue with me? ;-)
My motor, which is a new already BMW 1 series (so the cheapest range they do!), scans the road signs and alerts me if I'm doing the wrong speed. Not GPS and maps - it uses OCR to read the road signs, displays them on my dash and tells me if I'm speeding.
And multiple lanes is a problem already solved. Again, my motor (and many more) have lane detection. It knows what lane I'm in and tells me if I'm in the wrong one.
My point is that most of your objections are already covered technically speaking. Just need to join the dots and iron out some kinks.
Exciting times!
I wondered when that was going to hit the streets.
Again read my original post... ;-)
Now there's more to the tech than what you see.
Note: BMW opened up and office in Chicago, in the Boeing Building.
Now why is that? ;-)
Again, posted Anon for a reason.
None of mine do. Nor will any of them, ever. Too much to go wrong.
One wonders how, exactly, these obvious non-driver executives are planning to get rid of actual drivers and real cars. There are many of us, and we are legally allowed to drive on public roads here in the USofA, and I suspect we always will be.
One also wonders what will happen the first time a self-driving car kills somebody ... it's only a matter of time, after all.
@Jake - None of mine [change gear for me]. Nor will any of them, ever.
Hopefully the act of changing gear will be as old as feeding horses anyway - the electric motor will kill it off. But seriously - when did you last see an automatic gearbox fail?
@Jake - what will happen the first time a self-driving car kills somebody?
It will be an event which will be analysed thoroughly and the lesson(s) learnt will be taken on board by all other vehicles on the road, and *never* repeated.
The possibilities for carnage are *far* higher if you put tired, distracted, fallible, meatsacks in control of a lethal weapon.
@john Robson.
It will be an event which will be analysed thoroughly and the lesson(s) learnt will be taken on board by all other vehicles on the road, and *never* repeated.
You put way to much trust in corporations, how many times have they been caught covering up defects as it was cheaper to pay the odd legal battle.
@Lionel Baden, initially people will die from autonomous cars. Just like loads of people died during the early years of flight. But each time lessons are learned, safety devices/procedures/practises installed, and fewer people die next time round. Flying has used this method to make it one of the safest methods of transport.
Taking the attitude that a single death with an autonomous car means that the whole tech should be shutdown immediately is anti-progress. Progress requires death to proceed. Understanding that and realizing that everything has a risk and everything has costs and everything has benefits allows use to proceed to a safe world in the future. Maybe not immediately, but it will be the long term goal.
Corps might think that hiding the odd defect is cheaper but only in monetary terms, not in goodwill. When they do this practice and no one buys their product anymore they pretty much shoot themselves in the foot with this bean counter mentality. So corps quickly learn that it is actually cheaper and better to solve the problems.
"Progress requires death to proceed."
Really? In this day and age of Health & Safety not to mention litigation?
50 or more years ago I would totally agree. My argument is that if we had the Health & Safety regimes that are in place today back in the day, America would never have been visited (I won't say discovered, as it was already occupied). Ditto for all the 'new' countries such as Oz, NZ, South Africa to name 3. Of course the Romans were there way before the Europeans, eh!!
By the time the planners had written out a 'risk assessment form" the potential crew of the ship (which undoubtedly was the cheapest they could find) would have been press-ganged into the navy or died of boredom.
The bottom line for all these regulations seems to be that 'no-one should die - ever!!" Strange that so many do in the course of living - and comparatively few of them on the roads as opposed to, oh, let me see, illness, old age, had enough of this over-regulated world of ours.
Meh!
@Lionel Baden
Even if they did ignore the odd defect - that would still leave these vehicles far safer than the current generation of control systems.
I strongly suspect that the legislative framework would react fairly strongly in these (by definition rare) events.
The current control systems don't pay attention at all times, they don't look all around them, they do get tired and distracted. They generally overestimate their own ability (last survey I saw said that something like 85% of drivers considered themselves 'better than average').
In fact the current control systems kill multiple thousands of people each and every year - more than 5 people EVERY DAY in the UK alone.
Proper legal controls would help, like a willingness to revoke licences (even fairly short bans would be effective if they were regularly issued), but the majority of deaths aren't caused by malice.
The driverless car is coming, and coming soon - and we'll all be better off for it. Those who enjoy the act of driving can go and play on a track*, where you can *really* enjoy your driving.
* Race or off road.
I strongly suspect that the legislative framework would react fairly strongly in these (by definition rare) events.
Except you forget one thing. Legislation is nearly always 15 or more years behind technology and often much further. Consider the first reliable ABS systems appeared about 1970 yet legislation mandating its use didn't come for another 37 years. You have to remember that lots of folks die every day traffic accidents and one more or less isn't going to raise any hackles and you're likely to have as many luddites railing for the banishment of these autonomous mechanized dealers of death as you'll have techno-utopians saying the tech is great and has the potential to be far better than what we have if we only had more laws. Either way the pols have little incentive to do anything since viewed as a competition between the firmware controlled cars and the wetware controlled units, the firmware needs only achieve a draw to survive but needs to be several orders of magnitude better in order for any pol to even consider banning wetware controls.
Oh, notice that all it takes is one incident of someone hacking a car and using it to run down a dozen kids in a playground and its likely it all gets shelved for years until it's "proven safe" again. That's why white hats like the guys who found the hole in FCA's Uconnect are necessary and more important than most VPs. Notice that it's the emotional events that will drive any change, the autonomous car could be hacked and stolen by the truck load and the lawmakers won't bat an eyelash unless it's their car.
"Hopefully the act of changing gear will be as old as feeding horses anyway - the electric motor will kill it off."
Electricity will never kill off feeding horses. Nor will "automatic" transmissions or electric motors. Horses & other manually driven vehicles are here to stay.
"But seriously - when did you last see an automatic gearbox fail?"
A couple months ago. A friend's Lexus transmission fried. Was pretty comical, actually. The guy went off the deep-end, went on and on about how "that's impossible, it's a Lexus!" ... despite the fact that I watched him bottom out on the dips in the road west-bound at the corner of Chase Street and Austin Avenue here in Sonoma. He refused to listen to me when I pointed out the leaking fluid ... and then told the engine control system to stop complaining.
"The possibilities for carnage are *far* higher if you put tired, distracted, fallible, meatsacks in control of a lethal weapon."
Shirley you are discussing fast-so-called-food rather than automobiles in the above sentence?
"Electricity will never kill off feeding horses"
Huh - deliberate misreading?
"Horses & other manually driven vehicles are here to stay"
No doubt - but neither will be used as a mode of transport. They will both be sport/hobby instead. Which is fine. I have no desire to stop people having their fun - but they can do so in an environment which doesn't threaten the lives of others...
There is some overlap for other forms of transport - Walking and cycling are used both for sport and transport, but they pose very limited external risk.
"He refused to listen to me when I pointed out the leaking fluid ... and then told the engine control system to stop complaining."
You know what - I don't think anyone can blame that on the gearbox. That's user failure.
I've driven a mix of manual and automatic, petrol and deisel cars for longer than I care to think about. If you ignore the warnings then anything will break. One big advantage of electric drive is that there are far fewer things to go wrong (potentially down to one moving part per wheel) - and I think that the electric motor will kill the gearbox.
"Shirley you are discussing fast-so-called-food rather than automobiles in the above sentence?"
Don't call me Shirley. And no - people (including myself) are by far and away the weakest link in the system. Even the current generation of autonomous vehicles are better than humans (including me).
@John Robson: "when did you last see an automatic gearbox fail?"
Try buying a Chrysler-Fiat vehicle. Specifically in their Dodge line. If you have a Dodge vehicle with a transmission that last 50,000 miles with no problems, count yourself lucky.
If you fail to provide proper maintenance? Involved in an accident and the repair was done wrong?
It happens.
Your automatic transmission?
AWD ~135K miles you may start to see issues.
But yeah, I agree with your points... Just had to be a dick and point out that mechanical failures happen. Hence only the 10 year 100K warranties. ;-P
"Intake of nicotine, fat, alcohol, salt and sugar is far worse when it comes to mortality ... Do you want those banned in your lifetime, too?"
If my eating sugar, fat or salt sure would kill other people, then sure, does it?
But wait, smoking does kill other people, could that be the reason that smoking is illegal in many places?
This, 'but I enjoy driving, it's the one time in my life in which I feel I am in control, who cares how many people get killed?' whining, won't stop the ban, get used to the idea.
@jake - "Intake of nicotine, fat, alcohol, salt and sugar is far worse when it comes to mortality ... Do you want those banned in your lifetime, too?"
Nicotine doesn't - it's the asscociated tar I think...
But the big difference is that (except for certain aspects of alcohol consumption and smoking, which are already legislated against) these items don't kill other people.
But there are already plans to levy taxes on salt/sugar rich foods, and I support those.
"
Intake of nicotine, fat, alcohol, salt and sugar is far worse when it comes to mortality ... Do you want those banned in your lifetime, too?
"
If they presented significant risk to *other* people, then I quite possibly would support a ban. As it is however, I believe that people should be permitted to risk *their own* lives as much as they like.
"driving is a hobby that kills hundreds of thousands of people every year"
"Intake of nicotine, fat, alcohol, salt and sugar is far worse when it comes to mortality ... Do you want those banned in your lifetime, too?"
Intensive research has shown that there is only one cause of Death.
Life. The only way to eliminate Death is to eliminate Life.
Or... Y'know, we could just accept the fact that one day we'll all die, and spend more time enjoying life, rather than fretting over how to cheat Death.
Oh, and funny thing about fat, salt, and sugar. Eliminate those from your diet, and you'll die much quicker than you will from over indulging.
Mine don't either. But that is because I choose to buy manual transmissions. My last 4 cars were all manuals. My only sadness is that it is hard to find a nice sedan with a manual, and it is getting very hard to find a sports car with a manual. The way I look at it, a sports car with an automatic is like a bicycle without a seat: it is fundamentally wrong.
"Hah! Try to find a station wagon with a manual transmission."
I own a drivable, street-legal, insured, not-yet-restored, first registered in 1969, Oldsmobile Vistacruiser. 455ci (7.5L, ish), Hurst 4 speed manual. The numbers match.
The 8-track player works, too, but that's hardly impressive ;-)
"
One wonders how, exactly, these obvious non-driver executives are planning to get rid of actual drivers and real cars.
"
There has been no suggestion of doing such a thing. Technology will slowly take over more of the functions of the driver to make life a bit easier. You and other "purists" will still be permitted 100% manual control for a long time to come, though I wonder whether you also spurn ABS? How about automatic advance/retard, automatic chokes, fuel mixture adjustment, synchromesh, self-parking wipers? All things that once had to be controlled by the driver.
My prediction is that fully automated driving will start with the provision of special lanes on major motorways that may only be used by vehicles fitted with self-drive technology. The technology that would enable driverless cars to safely negotiate such a lane is already available. Much of the hazard detection would probably be external to the cars and part of the road infrastructure, with hazards being communicated to the vehicles on the affected section by some form of wireless signalling.
In the case of serious accidents there would be an investigation just as there is at present, and the investigation would determine the cause of the accident. If the cause was due to the automatic system, then the manufacturer would have to pay compensation, and if a systemic fault there would be a recall - the same as when an autopilot was the main cause of an aircraft crash. It is therefore in the financial interests of the manufacturer to ensure that failures are a rare event, and the safety record of automatic systems is bound to be better than that of human drivers (not a particularly high hurdle). That's not to say that there is not likely to be a huge public outcry to ban driverless cars after the first fatality, even if it was caused by a drunken pedestrian stepping directly in front of the automatic car, but as such cars become familiar people will become less frightened of them, just as few people these days fear that microwave ovens are dangerous even if someone occasionally tries to microwave something that explodes and kills them,
The smart money should be on delivering goods. not people!
- Over 50% of urban journeys are delivery
- You can't kill the occupant
- Goods are more patient than people
- Goods are more robust than people
so:-
An autonomous locomotive that pulls a series of "pallet trailers" holding standardised "boxes"
A network of "shunting yards" to rationalise journeys.
The loco can disassemble a pallet and deliver to the door
(A bit like Victorian trains really!)
Making the goods train, low axle weight and just big enough to carry a standard pallet, means that two would fit side-by-side in a EU standard road-lane, and Road damage obeys a fourth power of axle-weight law!
Whilst the step to fully autonomous everything is clearly a step to far for many to contemplate, it makes sense that all the small improvements - from vehicle-aware cruise control to self-parking are steadily changing our relationship with cars. You can understand that we're not just going to wake up one morning and hand the keys over to our smartphone, but the little conveniences are going to accumulate until we only need to take control of the car for the interesting bits.
Keen drivers might baulk at this, but we have to recognise that most car journeys are dull - commuting, school runs and supermarket trips. Who hasn't set off on an unusual journey and automatically turned left to go to work out of sheer habit? Between Uber, car sharing and cars that can deliver themselves, these boring journeys are ripe for handing over to the machines.
Now, the important question is - how much of this work is being done in the UK? We do some class-leading engineering in the automotive sector, but vehicle IT is a new niche and we're historically more interested in the greasy bits than the wiring. The integrated solutions that are going to be needed will likely have to be developed holistically, so there's a real possibility that we could get locked out of the industry. I'm sure there will be universities looking into this, but are there manufacturers out there ready to move this from theory to practice? It sounds like a great area to be involved in.
I think the tipping point for self-driving cars is not going to be when *people* accept them but when *insurance companies* accept them: If the *at-fault* accident rate of self-driving cars is 5% (an example %, a lot higher than Google car accident rates) of human drivers, then self-driving car owners are going to expect either a/ a massive discount or b/ manual driver insurance to rise a lot.
Once that happens then the average car owner will have a no-brainer decision - dirt cheap or sky high insurance.
"
It will change from insurance to product liability.
"
No more than it is at present. The person or company that will have to pay (and therefore will need insurance) will be the person or company that was to blame for the accident. In most cases the blame will *not* be with the self-driving car manufacturer. If the automatic system failed to avoid an accident that it should have been able to avoid, the cause of the failure will determine who pays, which might be the manufacturer if it was a design fault, but may well be the owner of the car who failed to maintain it properly, or a garage that failed to service or repair it correctly. If your wheel comes off and causes a fatal accident, it is not Dunlop or the car manufacturer who is held to blame *unless* there was a design or manufacturing fault in the wheel or fasteners. What we may well find is that the number of accidents caused by owners/drivers fall to such a low level that 3rd party insurance will no longer be mandatory - any more than pedestrians or cyclists have to have insurance even though they sometimes cause expensive accidents. It may be of interest to know that 3rd party insurance is not mandatory for private pilots in the UK for a similar reason.
Or you might have an airbnb-for-cars scheme where you drive to work, then the car spends its time in a car share, with people using apps to hire and drive it until it's time for you to leave. You drive home and then put it out on the street again. A new generation of pimp my ride.
The last thing I want when I leave work to sit in a traffic jam is my car smelling of burger king, puke and other peoples farts.
That's easy:
* You don't own the car but rent transport facilities. (Like city car clubs.)
* The cars have cameras inside.
* If a car arrives for you and it's unacceptably untidy, you report it, it goes away to get cleaned and you get another.
* The previous user gets charged for the excessive cleaning, through their normal account - and the cameras provide evidence.
Once driverless cars become common, I think that the car rental companies will merge with taxi companies and the majority of current car owners doing less than say 10K miles per year will cease to own their own car and simply rent a driverless one on a per journey basis.
Ford are more than capable of selling direct to fleet buyers. It's the dealerships that will have no one to sell to.
I can't wait to get back in an auto. Constantly changing gear manually in traffic is simply a complete waste of effort. An auto box can last hundreds of thousands of miles with proper servicing - fluid replacement, filters if possible.
There would be no need to connect said gearbox to anything external to the car. Auto boxes have been around for decades without needing any form of outside control.
What worries me FAR more is the current Ford advert for park assist showing a car steering itself.
NO NO NO AND AGAIN NO!!!!
" An auto box can last hundreds of thousands of miles with proper servicing - fluid replacement, filters if possible."
Yes, it can last a long time. And it can also fail in less than 50K miles, even with proper servicing. And when they fail, it can cost alot - $3K to $5K.
"Constantly changing gear manually in traffic is simply a complete waste of effort."
Maybe so, but it probably keeps your mind on the job more than an auto does.
I keep having the feeling that the more gizmos added to 'help' the driver is in fact ensuring that the driver pays even less attention to what they are doing.
Tell me how a touch screen which has to be looked at to operate is safer than manual knobs and buttons which can be used mostly without looking.
I'll take the electronics to give better fuel economy. I'll even take ABS but I refuse to entertain a car which starts telling me how I should drive. I have an older car with a Tiptronic gearbox which I use as a manual. The only reason I have an auto is that in my ageing years (ha ha) if something should happen to me, my wife can drive it. I have never found an auto gearbox that changes gear when I want it changed, and that's either changing up OR down, which no-one seems to do these days. All you get are brake lights flashing all the time which is distracting especially on a wet night. One exception was a Renault Laguna which would automatically change down a gear if it detected (presumably) that you were gaining speed but no throttle input. CVT works better on that score as well.
"Constantly changing gear manually in traffic is simply a complete waste of effort."
Incidentally, the Renault gearbox crapped out on me rather spectacularly when it decided to drop into 'limp home' mode (2nd gear) at 70 mph in the outside lane of the M3.
The effect was as though I had completey lost power, and made frightful noise from the engine compartment, and a definite moistening of the nether regions, not to mention frantic signalling and looking for a gap in the traffic in order to cross 3 lanes of traffic at a much reduced speed. I discovered I had enough driveabillity to get home, fortunately not that far, after pulling to the side of the road.
It sounded expensive - and was! Around 1000 quid for a replacement, and that was from a wrecker.
Turned out to be the result of a hairline crack in the casing causing a pressure drop and a sensor triggering the limp home mode.
I wonder how an autonomous car would have handled that little situation. Hmmm?
I love driving. I love the control, I love the "game" of optimising my path and evaluating what's going on in front of me to predict what's coming next. Not to mention the visceral feel of a reasonably sporty car (or my bike) on an interesting road.
I'd take an automated car in a heartbeat for a daily driver. With a young child I have very little time to myself, and if I had 60 minutes a day to read or listen to music undistracted it'd be worth its weight in printer ink for me.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
