Audi's self-parking car: What's stopping the tech getting on our roads?
Summary: Audi recently demonstrated a car that can park itself without the need for a driver. But while the technology already exists to make autonomous cars a reality, it will be a long time before our vehicles will take over the driving.
Car manufacturers aren't shy of showcasing their forthcoming premium features at trade shows like last week's CES or this week's Detroit Auto Show. It's especially true of Audi, the German car maker whose slogan 'Vorsprung durch Technik' translates into 'Leap ahead through technology'.
For Audi, one technology with which it hopes to leap ahead is a local Wi-Fi hotspot and integrated 3G/4G for internet access for passengers while on the road.
But the most impressive technology-led iniative for the company is still a way off: a self-parking car, or, as Audi calls it, Piloted Parking.
The idea behind it is rather simple: you exit your vehicle, pull out your smartphone, click a button and the car drives itself off to a parking spot at a nearby garage. What at first sounds like a mixture between Knight Rider's KITT and something James Bond's Q might create is actually technically possible. Audi demonstrated the feature last week at CES in a secluded area of a Las Vegas hotel.

The car uses an array of internal and external sensors to get its position: Audi claims they can be as accurate up to 10cm, but only if they have access to special laser sensors inside the parking structure (four of those scanners had been set up in the parking structure to support the demo). These might be redundant in the future, as the car maker is working on a laser sensor that will be integrated in the car itself (think the sensor tower on top of Google's self-driving car, but completely integrated in the chassis).
The self-parking system also needs access to the car park's management system, in order to find and allocate a free parking space and transmit the route to the car. Since most modern car parks have more than one level or are underground, GPS-based positioning is not really an option, so instead the management system uses Wi-Fi to transmit the route.

The legislation roadblock
Cars are obviously capable of finding their own way around already, so what keeps the autonomous automobile from completely taking over our streets? Simple: The 19 Convention on Road Traffic, drafted in Vienna in 1968 (PDF), which doubles as a UN treaty. Article 8 of the treaty contains the vital point: "Every driver shall at all times be able to control his vehicle or to guide his animals". Countries like Germany or Austria ratified the agreement and incorporated it into their local laws. Other countries, such as the UK or Japan, signed the treaty but didn't ratify it completely. The US didn't sign the treaty at all, however - making it possible to drive cars autonomously in California and Nevada.
Another big problem is liability. Who is responsible if an autonomous car is involved in an accident? The owner, the manufacturer or someone else? While Google's self-driving cars were able to complete more than 300,000 miles without an accident, drivers and insurers still need to plan for the possibility.

The third issue comes from the drivers themselves. Imagine you've invested in a premium car, including a nice, powerful engine, and then that car tells you to relax, it will take over – not all drivers would welcome ceding control of their vehicle to the vehicle itself.
Audi's solution to that is rather simple: "The car should only take over if the driver is OK with it or if the current driving situation is an annoying task. That's why we're focusing our autonomous driving scenario on situations like parking or traffic jams. When I want to have fun, I drive myself," Ricky Hudi, Audi's chief executive engineer, electrics and electronics told ZDNet. Another Audi spokesman agreed: "We need to build up confidence. The driver should be able to trust the car and to delegate tasks that bore him or her."
Despite the problems facing self-driving cars, Hudi is confident that they will be a reality in this decade. Early adoption will be in countries like Japan, where the growing traffic is presenting a problem. The replacement cycle of cars has its part to play too: with drivers only replacing their cars every seven to 10 years, it may be some time before self-driven cars are a common sight on our roads - even if the technology is ready to leave the garage now.
Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.
Talkback
Not for me, thank you
Having said that however, for persons with certain disabilities, I can see such systems being of considerable assistance. Some people simply cannot drive safely, but still need/deserve the independence and convenience that "driving" can bring.
Can I have your computer?
I want one.
Comprehension?
Replacing a car's "auto parking" capabilities with my own is trivial, simpler and cheaper.
Replacing my computers capabilities is none of the above.
If your post were an attempt at humor, it failed imho, but no offense taken. If it were an attempt at ridicule, you only ridiculed yourself by posting an idiotic and irrelevant response.
Take your pick.
Actually, the guy above you is correct, and you didn't sound like you love
That sounds like the type of things that were said when the automobile first came around (the horseless carriage), and the first computers first came on the scene. People resisted change, and change happened anyway, and those refuseniks eventually came around to accept the changes. Eventually, the self-driving care will take over, and you will have been assimilated. Plus, the benefits and major savings will being most people around to accepting the technology.
Sorry adornoe, but this comment of yours
is Bull, imo. First off, I don't believe that they will ever force anyone to give up having the choice wether they can drive their vehicle or not, at least not in my lifetime!
What about people with antique autos and motorcycles, etc.. Not gonna happen.
And how do you know, or where is the proof that it will be cheaper (major savings)? Just asking.
Benefits? Ok, I can agree to that, as it could benefit some people, but I have a hard time it's going to benefit everyone. What happens when the car is self driving, and a deer comes out into your lane and another car is coming the other way?
Is it going to make the right choice, and take out the deer rather than a family in another car, or is it going to swerve to the car in the logic that the other car will also be self-driving and will swerve out of the way also?
What happens when the other car doesn't have it? Or if it does choose the deer, will it then adjust it's programming to always go after a deer when it sense's one?? j/k
Sorry, couldn't help myself, long work night, though having a truck that huints deer could certainly help out some soup kitchens, just have to pick thru and choose the good meat..LOL
TW
Big difference...
Personally I don't think we will see cars driving themselves for at least 30 years. Why so long? Because we don't know the ramifications. The fact that Google has a car that drove around by itself is actually silly.
Think about it, when the first car came out how many accidents did it have? I am not talking Model T Ford times. I am talking before that. When people were still riding horses. During those days accidents were actually pretty rare since well, cars drove really slowly and there were no other cars around.
The moment multiple computers are involved we will have a clusterf**k of accidents. Google Airbus and its first set of attempts to reduce the pilot to a game console playing "helper". There were several accidents, and still are because pilot and computer "debate" on what appropriate action should be. Now take that and multiply it by a factor of a million. NOT PRETTY!!!! Until computers can reason, and make reasonable tradeoffs, they should never be allowed to drive.
I Totally Agree ... No Thanks (nt)
Definitely for me
There has been a lot research to get to the point where a car can actually drive itself, but driving mass-market adoption of the technology will be even harder than making it work in the first place. If this takes us one step closer towards acceptance, I'm all for it.
We already have "self-driving" vehicles...
And just who is going to be liable when one of these things has an accident while in self-driving mode? The owner? The auto manufacturer? The software company who wrote the code? The programmer who installed it?
One thing is for sure...a lot of lawyers will have jobs for decades to come, if this crap becomes mainstream.
Actually, with the self-driven cars, the losses will be a lot fewer,
Oh the idealist...
Here is a very very simple example. You have a cross with four lights. It is red for the driver A, but his car for whatever reason skids across the crossing. A car with driver B is coming the other way, and for them its green. What does the computer in driver B do?
Do they brake? Problem with that is that there are cars behind driver B, which will cause an accident on driver B. In this case driver A would skid across without accident.
Do they get hit? In this case driver A hits driver B, but the drivers behind are able to stop in time.
Do they brake and hit at the same time? In this case its worst case scenario and there are a whole bunch of accidents.
You really think a computer can handle this situation better than a human? I would reply, not a chance!!!
What would be better are automation to help a human make better decisions. For example ABS, or traction control. Or how about the front and rear sensor controls? Since I have these additions I have not scrapped my car at all. BTW in Europe the roads and parking lots are much tighter and prone to scrapping.
Oh, the skeptics...
Mind you that, a computer in control of a car, would be able to make millions of decisions in a second, while a car with a driver, would still be hampered by a human who cannot look in all directions at once, and has a reaction time thousands of times slower than a computer controlled car.
The ideal situation won't be reached until all cars on the road are all driverless, but, a driverless car will also be equipped to know when the car in front or in back or on either side, is being controlled by a human, thus, being able to compensate for the error prone human at the wheel. A human would be less predictable, and therefore, the driverless car might compensate by giving the human the right-of-way.
Driverless cars will come equipped with communications capabilities which allow each car to communicate will all other cars in the vicinity, and all of them could arrive at a decision which takes in to consideration road conditions and tightness and crossroads and lights and people and other objects in the surroundings.
It's not a simple proposition, where it's about, who gets the right-of-way, or who gets to a parking space first. It's much more complicated than that, and much more efficient than you imagine.
Real world
Unless these smart systems can learn to deal with unexpected, irrational, idiotic, unpredictable and sudden darting movements by human beings in unexpected situations and places - they are really not usable imho.
Googles Cars
I think using sensors is a pretty good thing: They don't get tired or distracted.
Where did you get the B.S.???
All the testing was done in an empty parking lot.
BTW, official records show at least 7 accidents (all minor) since they started testing.
I got it from
Can we trust these results?
In short someone will make a lot of money.
Unfortunatly also more than enough reason for slanting test results. The results seem too good to be true.
The kind of results Google claimes need independent verification. Based on the results I suspect that something is wrong with the experiment.
If there never was a situation that the automation couldn't handle the sample is either biased or too small. If the driver was there to take the controls when such a situation happened then biased might be too generous a word. Having a human backup in such an experiment is reasonable. Claiming that the accident only happend because the human was driving would not.
All this is speculation. Has anything been published that provideds enough detail to judge the validity of this testing.
Google won't be able to "slant" test results, because, before any
The tests are just that, tests, and not a license to be used with no regulations and no independent tests. Once the larger set of tests are conducted by government and insurances companies and independent consumer agencies, then, and only then, would the figures coming out of the tests be deemed valid.
Well.....
I realize I may to some extent be arguing against my first post, but I do not consider myself "stupid", "careless", "irrational", "idiotic", nor "unpredictable". I do not disagree with that general characterization however.