Saturday 16th May 2015
How long before we see self-driving cars on roads ?
The news that
Google have this week been given the green light to test drive their driver-less cars on public roads is pretty interesting.
Initially those driverless / self-driving / robot cars will be tested on public roads in Mountain View, California, where Google is.
As you will see in the article, no-one expects them in general circulation for at least.......another 5 years !
Well, personally, I think that, in the western world, with its strong rule of law, these cars will never drive on public roads in general circulation other than in very regimented environments.
Here are a few reasons why :
Intrusion of privacy : In order to prevent accidents
and, critically, to assist accident claims or defend accident claims, these cars are full of cameras recording everything as they go along (they also use laser radars and GPS sensors). As soon as those cars are on the road it won't be long before someone sues Google or the owners of such cars for intrusion into their private lives. And that's just in the US
— in many countries in Europe, privacy laws are even stronger than in the States. And what about me owning one of those cars ? Will I have access to all the video footage to assist me ? If I do, what if I see on the video footage the private life of the first floor of 37 Johnson Avenue ? Potentially even worse, what if I my car is the witness of a gang or Mafia hit, or, even, an FBI hit ? As the owner of the witness at the scene, will I be made to..... disappear ? And where will all this footage be kept
— after all, someone may claim two weeks later, two months later, a year later.....
Legal claims : In the 8 months since September 2014, when the reporting of accidents was a condition of Google eventually being allowed to test its cars on public roads, Google's driver-less cars have been involved in 5 accidents. Google says that none of the accidents were its fault, but it has not made public the record of the accidents and the California Department of Motor Vehicles says, quite correctly, that it cannot disclose details from accident reports. 5 accidents in 8 months from 50 cars.
So, while limited to 25mph, and not allowed on public roads and the craziness of city driving, 10% of Google's cars have had an accident every two months. Now, Google tell us their cars did not cause the accidents, but that is not how it works in the real world : you discuss it between the two drivers, and, if you disagree, the police gets involved. In many countries, like my own, France, the police has to be involved regardless. So, when you have an accident with a driverless car, who do you discuss the accident with ? The passengers ? Who do you claim against, the owner or the car maker ? Do you see my drift ? Now, if you are a passenger in a driver-less car in an accident, why not sue the maker, Google, regardless of whose fault it was.... Why not ! Or are these cars going to have the passengers under round-the-clock surveillance as well ? And, while we're at it, what about the witnesses of the accident : initially many will have, on the spur of the moment, such a dislike of the driver-less car that you can't talk to, that they will falsify their witness statements. It will become such a legal such a mess that some cities will start considering putting street cameras on specific highways (more intrusion of privacy) and driver-less cars will not be able to stray off those highways because, as we all know from all those YouTube videos of Russian car crashes or helmet-camera-wearing cyclists, the camera on the car does not give the whole picture, it only gives the picture from the driver-less car's point of view ! I could go on
— the more one thinks of it, the only picture that comes out is an ever increasing legal mess. And, trust me, insurance companies do not like legal messes and, in the end, they will kill off the driver-less car because there will be so many challenges and court cases.
Technical Problems : If the above legal issues are not enough to scare off law makers into not allowing driver-less cars other than under extremely regimented conditions, then consider this :
2014 was a record year for the number of car safety recalls by the automotive industry. In the first 4.5 months of this year alone, 2015 is on course for beating last year by mid-year ! Try
"Exploding airbags which can kill their passengers, and have killed some" as the cause of car recalls
— 24 millions in the two years to December 2014, and just this week, on 15-May-2015, Honda is recalling 4.9 million cars for their Takata-made airbags. It is a fact of life, manufacturers make mistakes, most of the time
oversight mistakes rather than plain bad manufacturing, but mistakes all the same. Despite all the testing Google is currently undertaking, imagine the impact of oversight mistake in self-driving cars : while in normal cars a recall is, effectively, left to the owners to bring their cars to the garage, in the case of driver-less cars it is almost certain that the result will be like the airplane industry, all driver-less cars will be mandatorily "grounded". Can you see the beginning of another legal nightmare for someone whose livelihood depends on their robot car....?
Technology that is too perfect : There have already been reports that those driver-less cars need infinitely much shorter braking distances to the car in front because they are so...... perfect ! So, you are driving along the road and a dog runs out in front of you. By the time that you, as a human, have reacted, you have hit the dog. You still apply the brakes, of course, but the dog has had it. In a driver-less car, however, the computers will react a thousand times quicker than a human, and will stop the car on a penny. The stop will be so instant the passengers will suffer injuries while the following driver, who had carefully left the recommended braking space between his car and the driver-less car, will still plough straight into the driver-less car because his/her
human reactions is so much slower than the driver-less car's reaction. Sort out this legal mess. After you have interviewed the driver-less car, of course ! Now, I hear some of you saying
"Yes, but Google will have thought of that and introduced progressive braking for such situations". Fine
— replace the dog with a human stepping out in front of the car. What happens now ? Does your robot car kill the human in favour of not injuring its passengers ?
And, how about when technology is not so perfect : it is
–15C outside, the roads have been gritted, the car ices up, the glass on the cameras, laser, and GPS sensors either ices up or fogs up. Hmmm....
A clarification : For the readers who think that I am saying robot vehicles do not and will not work, I am of course not saying that. We have robot / self-driving vehicles everywhere in a lot of places nowadays, and in Australian mines they have gigantic trucks the size of small planes that are entirely driver-less, carrying tons of minerals from the mines to the processing areas, but in all these instances, the environment is completely controlled, completely the opposite of our roads and cities where humans, and animals, do crazy things.
Conclusion : when all these scenarios start playing themselves out in front of us and make the front pages, many insurance companies will refuse to insure driver-less cars, and many cities will ban them, and many other cities will decide, instead, to build good old fashioned tramways, more subway stations, etc.., rather than have driver-less cars.
You've read it
I am off to the
bookmakers to put a bet that, in 2020, I will have been proved right ! ☺
—— (TUT) SpaceMan