Googleís recent announcement that it would produce approximately 100 self-driving cars came as a surprise to the general public, but itís hardly a new idea within the confines of the auto industry.
Enthusiasm for autonomous vehicles has grown in the past decade, as automakers and universities have teamed to create cars capable of steering, braking, accelerating, and navigating on their own. Starting in 2004, races sponsored the Defense Advanced Research Projects Agency (DARPA) proved that a self-driven car was a viable concept, with five vehicles completing a rugged 132-mile course in the second race and six more finishing the third. Since that time, Google has logged more than 700,000 miles on its self-driving cars.
Here, we present a short history of autonomous cars. From student-designed buggies to self-driving production cars, we offer a glimpse at the future of the automobile.
Click on Chevroletís EN-V below to start the slideshow.
Could this be the future of the automobile? Jointly developed with Segway Inc., Chevroletís EN-V (Electric Networked-Vehicle) is a self-driven, electrically propelled concept car designed for urban environments. Chevy said the autonomous two-seat vehicle is designed to address environmental issues, traffic congestion, parking, safety, and energy consumption. It will reportedly be featured in a 2015 Disney film called Tomorrowland. (Source: Chevrolet)
I agree with you, Pudubu. A road full of autonomous cars would definitely reduce the annual 30,000 highway fatalities in this country. On the other hand, I'm not confident that automakers will be willing to do this, due to legal fears. Look how much Toyota lost in the unintended acceleration case. If automakers have to set aside billions of dollars for legal costs, it's going to make them think twice about this, no matter how many lives might be saved in the process.
Unfortunately while the only solutions require active sensors I don't really see this becoming a reality. Headlights->object->human works as long as there are no oncoming headlights washing out the reflected signal. If we have hundreds of cars with lidars the anti-jamming that is going to be required will make it infeasable.
Don't get me wrong, I spend 1.5hrs a day actively driving and could probably run a YouTube channel with the crazy stuff I see everyday. You will be hard pressed to find someone more keen on getting the meatbag out of the driver's seat.
The hard part, as with most sensible things, will be convincing law makers, because lets face it the control freaks in parliament /congress are not going to go for a machine driving them around for at least 2 generations.
LIDAR is a little more sophisticated than headlights. It is more like the situation we have with cell phones, where each signal includes coding such that the source can differentiate its signals from the 100's of others propagating through the same space.
See the posting by "Ratsky" in the comments on "What's That Thing Atop the Google Car?"
Chuck, that is the rub. The legal liability is going to overwhelm this.
The solution that makes sense is a system that uses sensors in the road, something like your second slide. In that case, the liability is spread between the government and the car maker. On the other hand, the start-up cost is outrageous. It may be worth it, but it will take a real act of political will.
Autonomous cars mixing it up with irrational-driver conventional vehicles makes for a far more complicated environment - insurance and legal environment, that is - than a lane set aside for autonomous only vehicles. Yes, these shopping trolleys will need their own lanes to keep the lawyers unemployed.
The legal issue, Pudubu, is essentially that automakers could be responsible for any accidents or fatalities, unless they can prove they've tested for every possibility, which is impossible. See link below.
I have a hard time with the whole concept...as much as I like the idea. I just think back, and not long ago, where Toyota had braking and I believe gas pedal issues. I just think it has a very very long way to go before I trust an auto car. If they can't get the basics right....I mean really. And Toyota has been doing it a lot longer than google.
I love the fact that the google cars have no steering wheel or pedals. So when it screws up I guess off the cliff you go! No way would I trust a car that I couldn't at least correct at some point. Well, at least it only goes 25mph, you'll probably survive if you have to bail!
The Toyota case is definitely a concern, Cadman-LT, especially when you place it in the context of the autonomous car. Software expert Michael Barr, whose testimony turned one of the recent Toyota trials, had a lot to say on that subject at EE Live this year:
Multiple reasons exist as to why the autonomous car willnever work. Not that they won't be made and sold, but that the promised benefits will never happen. First is the legal problem, since when there is an accident with injury somebody will be at fault, and that would be the producer of the control program that was doing the driving. The auto companies won't accept that, as Charles has already explained. And he is right. The second challenge will pop up even sooner, which is that when an exception situation happens the car will not know how to deal with it and so it will just stop. And we know that exceptions will happen. That will make them very unpopular in a real hurry. Or they will make a wrong choice,with a similar result. The automotive equivalent of the dreadd "blue screen of death" will be te end of the driverless vehicles on public roads.
Charles, I have seen what the "BSOD" can do to a computer contrllled industrial testing machine, (NOT of my design), and the damage was not minor.
The comments about updates and hackers are also certainly valid, and they do bring up whole additional realms of reasons to approach the implementation with caution.
My prediction is that if the autonomous vehicles are ever mandated that traffic speeds will be reduced to the point that any collision would not cause injury, since the vehicles will be moving so slowly. While that mode could be safer, I don't expect that many would find it acceptable. The entire exercise will turn out to have been a large waste of money, materials, and time. And only 80% of the folks will realize it.
A zero risk situation winds up being one in which no choice can lead to any injury. That sounds a lot like the situation of dead people. Any life worth living includes making decisions, and wrong choices can lead to problems. The zero risk autos will be so slow that they will be completely worthless. Any zero risk control software will be in a constant state of panic because some exception may occur, and by definition it is not prepared to handle any exceptions.
Reducing traffic injuries would take a few simple but drastic changes: 1. Only license people willing and able tp focus their atention on driving. 2. Remove the distractions from the drivers situation. This would mean a minimal number of controls. 3. Test reaction times at each license renewal, and don't renew those with the excess response delays.4. As part of the license test require an understanding of some vehicl kinematics, such as braking ability and turning dynamics. 5. Do not issue a license to anybody who can't read.
For starters this would cut the traffic on the roadways in half, which would reduce all kinds of problems. And those drivers remaining would be much safer drivers. PLUS, public transportation would suddenly have enough riders to be profitable.
I think the march of history and an aging population will make this inevitable. Recall that in the early 20th century automobiles themselves were the subject of great concerns and restrictive legislation (they scare the horses!). When my eyesight, hearing, and reflexes degrade sufficiently, I will be happy to turn over driving duties to Google.
Thanks Charles for such an interesting post , No doubt autonomous vehicles are becomming very in these days and they are also becomming the necessaity as well because it results in much safer journey, and because of sensors these cars will be able to pack closer together allowing more cars on the road, It will avoid the hassle of learning driving and then getting the lisence , According to me it will make the lives of passengers easier in the area where there is parking issues as well.
With every technology there are pros and cons associated no doubt autonomous cars have a lot of advantages but definitely there are cetain disadvantages also associated with it like there will always be a worry of malfunctioning of computer or system, It will make many people unemployeed for example taxi drivers, truck drivers etc , Initally this technology will be very expensive so every one wont be able to afford it .
Also, entrance and exit to the EN-V is just like the old BMW Isetta. A front door refrigerator. It was deemed too dangerous then, it is still too dangerous. A frontal accident would trap the occupents.
My biggest concern on these vehicles are that they are all controlled by computer. And just like my Windows at work, we will be getting inundated with updates, and the computer part of the car will bog down. My GPS has more than once sent me to the wrong address, or onto the wrong way of a one way road.
Then, because we are linked to Google, we will be subject to Hacks. What mischief could a hacker do? 'nuff said.
In actualilty, there are many situations where automatic driving would be preferred. Especially, long drives on the expressway. I would prefer to see railroads opened up to continuous use traffic. We can just drive our (electric) car with minimal range onto a personal rail car (like a ferry), and program in our destination. This would also work for towing camping trailers, or any long distance trucking.
Very impressive slide deck of autonomous vehicles. I found the MI-Detector to be quite interesting because the box at first glance looks like a geiger counter. It's amazing to go back into the archives and see how today's sophisticated technologies are influenced or originated using simple solid state electronics. Again, very nice slide decks.
The question of whether engineers could have foreseen the shortcut maintenance procedures that led to the crash of American Airlines Flight 191 in 1979 will probably linger for as long as there is an engineering profession.
More than 35 years later, the post-mortem on one of the countryís worst engineering disasters appears to be simple. A contractor asked for a change in an original design. The change was approved by engineers, later resulting in a mammoth structural collapse that killed 114 people and injured 216 more.
If youíre an embedded systems engineer whose analog capabilities are getting a little bit rusty, then youíll want to take note of an upcoming Design News Continuing Education Center class, ďAnalog Design for the Digital World,Ē running Monday, Nov. 17 through Friday, Nov. 21.
Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.