Those dates seem both far away and close, Chuck. Since Google has already developed one, a couple years ago, and since it was able to drive -- driverless -- from San Francisco to Los Angeles, why will it take so long?
I believe that we will be able to experience the fully automated vehicles in near future. Why I am saying so is recently launched Mercedes Benz S-Class which is the top end of that segment has this option. Its need to have only two fingers on the steering, that is also because of the legal purposes otherwise it will drive automatically.
Pubudu, I agree, It's clear that the technology is there. I wonder what the 20-year holdup is? Is it acceptance by cities and states? Acceptance by vehicle buyers? I don't think it's the technology that's the holdup.
I think the real issue with making self-driving cars is they will have to compete with driver-operated cars. How much of our highways will have to be reserved for autonomous vehicles? The technology exists today to make a fully autonomous vehicle, as is demonstrated in the various DARPA challenges. There will never be technology sufficient to adequately predict human behavior. Drivers will need more and better skills to deal with the occasional failure of the technology, however with less and less practice, their skills will likely deteriorate. I really think if drivers wish to spend their commute time more productively they should consider public transportation.
Bob you have made a very valid point. Transition to a new technology is always the biggest hurdle. Self-driving cars do exist. But a proper planning needs to be done on how to introduce them on roads. The transition should such that you can safely use the new technology with the old one. Like when internet introduced the IPv6 address, they had to find a way to collaborate it with IPv4, without changing the basic network structure upon which it was operating.
I think the 20 year hold up is to improve the tech to make it foolproof. Even though the driver is to pay attention in case the automation fails, we all know they will disingage and either nap, have their nose in a newspaper (er, e-reader), watching a movie, or play a game.
Emergency reaction times will skyrocket and our litiguous society will blame the manufacturer instead of themselves. Why do you think irons have tags telling you not to use them on clothes you're wearing?
Pubudu, thats really great just two fingers on stearing just for legal purposes and the car will drive you strange and impressive . Like Is this Mercedes in market available and what about the peddal does it have a pedal or its completely pedal free.
self driving cars will never be as efficient as a collection of good drivers driving in areas designed to allow efficient traffic flow. Please note those qualifiers, because they describe a large portion of the reason that any motivation for a self driving car exists. Way to many roadways and streets are set up to impede the flow of traffic, and a whole lot of drivers are either not able to concentrate enough to drive safely, or they are unwilling to pay adequate attention to driving in a safe manner. In addition there is a portion of the public who are so chained to their fears that while they can drive safely they will not be able to do it at anywhere near the speeds that are efficient.
Dealing with drunk and impaired drivers is a separate issue, and they will be an even greater hazard with automated cars because their actions will not be normal or logical.
And I still offer the question about who will be liable in the event that some system failure causes an accident resulting in damage and liability. After all, cars are much less reliable than airplanes and missiles.
And one more question: Would you be willing to ride in a vehicle that had a control computer with a microsoft operating system?
"self driving cars will never be as efficient as a collection of good drivers"
You obviously have not driven in the Chicago area, where the roads are great, but anything on the side of the road (either side) will create logjams from people slowing down to rubberneck and once that happens, it takes forever to clear and the affected band of traffic will migrate miles away from where the problem was and have you wondering what triggered it.
An automated system will not be distracted, may even play nice when it comes to merging. I used to fantasize about selling bumperstickers that said "DEATH BEFORE MERGING" since it seemed to be the mindset there.
Brainiac, you will note that I did use the word "good", and it was intended to be taken in it's true meaning. Stupid gawkers will always cause a traffic slowdown, and the only solution is a serious ticket for intentionally creating a traffic hazard. Two or three hits of $500 should help reduce the symptom, although the contrary is also true, that "you can't fix stupid". But still, reasojnably good drivers are usually able to handle most exceptions, while the very best that an autonomus car could do would be to apply the brakes. And I still ask what that driverless car will do when a truck passing by causes a large piece of cardboard to suddenly fly up off the pavement. It happened in front of me once and I ran over the cardboard because there was no other option at the time. All three lanes full at 55MPH and minimum headway because it was an open stretch of road and we were all going the limit. A robot car slamming on the brakes would have caused a large disturbance and probably a number of collisions. Of course a panicy driver might have done that also.
By the way, from what I have seen Chicago has more really poor drivers than Detroit has. Plus totally expensive toll roads.
Hahaha 55? That's what the signs say, but you better be doing 70 to survive in Chicago. (I lived there for 25 years) Like you, I was initially disgusted with the tollways, but then discovered they were the only roads actively maintained and improved. Also they've switched to open road tolling with transponders so you no longer have to stop. I was there during the transition to it as an early adoptor and it was fantastic.
I will say that learning to drive in that environment made me a better driver and the one good thing I can say about Chicago drivers is that they use and pay attention to turn signals. I used to carpool with a coworker who said never to use them in Dallas/Fort Worth where he had lived, because it only signalled your intention to the enemy who would use it to purposely cut you off. (I'm only repeating what I was told)
But good luck trying to give anyone a ticket. They only do that when you go above 80, otherwise that would cause its own logjam from rubbernecking and they want to keep things moving during rush hours.
The idea with getting all the cars automated, would be that they would all react to the emergency, cascading back from the incident. It would only be the inattentive manual driver that would cause an accident.
Brainiac, I only pass through the Chicago area a few times a year and we have checked the transponders and it would not be economical for us to use one. An alternate choice would be some means to render a photo of our license tag unreadable, witout the action being obvious to anybody watching. Illeagal of course, but an interesting concept.
I agree, TJ - the advantage of being able to spend the time commuting in other endeavors is very enticing. I think we will also be seeing cars available with additional accessories that enhance the ability to multitask once the role of driver is relinquished to the car itself. Sometimes it feels like I am a kid again watching The Jetsons when I read stuff like this...
Rob, I still have to wonder how close these things are. With the sensor systems we have today, coupled with the processing power, I can accept that these things will happen in the foreseeable future. On the other hand (here it comes) these things have been worked on for over twenty years (more like 30) by the military and aerospace industry. So, who knows.
I think the goal should be semi-autonomous vehicles for now. Let the computer drive on the highway, for example. That would make the roads more efficient (they could be coordinated via a central system, for example) and is probably easier than situations where you have to account for pedestrians. Once that is achieved, then we can look at where we go from there. Self driving farm tractors are already here. They use a different technique of processing the GPS signal that gives a much higher accuracy. So, that aspect is already well understood and used.
I will just leave you with a story from a friend of me who was a jet fighter pilot in the 1960s and 1970s. He was testing the first nap of the earth systems in a Phantom II jet. The pilot could easily grab the stick and take back control. This was done so that he could respond if there was a problem. My friend had a very hard time resisting the urge. I think two things were happening. First, he was skeptical of the whole idea. Second, he was used to being in control. As he put it, it was sometimes frightening.
I agree with everything you said, Naperlou -- except, Google has already created a self-driving car. It was about three years ago. After testing local trips, the company finally did a test where the vehicle drove from the Google headquarters in the Bay Area down to the Google office in Los Angeles. That pretty much says it all.
I think confidence levels would be higher if all the vehicles on the road were autonomous, naperlou. One of the fears is that driverless cars will get confused by the crazy things that humans do -- such as blowing red lights, not stopping before a right turn on red, or not waiting their turn at a stop sign.
Thanks Charles for such an informative post, but one point which i was wondering was that in future cars will be developed which will take off while the drivers stuck in traffic jam i thought a lot about it but couldnt understand what does this mean and how will the car take off during traffic jam. Will it take off like a plane or what ?
@Naperlou these things or technologies are not far away . We in 2008 developed a prototype of a completely autonomous vehicle which could drive you to the particular position with the help of calculations of latitude and longitude done by GPS. And the locations were inputted with the help of SMS . AT commands were used , Secondly sensors were used for obstacle detection . To cut the short story down when we as students developed the prototype its not that difficult for such big organisations to develop and launch them.
Debera, we already have the technology. As I mentioned, this is stuff the military has been working on, and solving, for some time. When I say we already have this, I am talking about farm equipment and some manufacturing systems.
I was talking to one farmer and admiring his big new tractor. There is a module for the tractor that will drive it completely automatically and dispesnse the seed (or fertilizer or perticide or herbicide) on a very fine grid. They use a different form of GPS processing that gets them inches of accuracy. What is interesting is that a human has to be in the cab to engage the tool. This is purely for liability reasons.
So, I fully accept that you, as students, could do this. I have seen some very good projects as a judge at various institutions. There is a lot more involved, though. It gets much more complicated when you have human drivers involved. If you had ONLY automated vehicles, then it becomes simpler.
That's a really good question, Rob. It seems like 20 years is quite a long time considering the technology already has been developed and proven. However, you're right that it also seems close. I personally am a little wary of this technology, although it sounds a dream to have your car drive you everywhere. But I also love to drive! So don't want to give that up too soon.
Yes, Elizabeth, I love to drive as well. Just finished a 400-mile round trip to bring my daughter home from college for Thanksgiving. And I enjoyed it. But with the crazy drivers on the road, I'm convinced self-driving cars will make the roads safer.
You could be right about self-driving cars, Rob, but only if the technology works as it should. Imagine a bug in a self-driving car. It doesn't just mean a hassle for the person who owns it, but it also could mean his/her life! But it seems the technology available already is fairly sound. It just needs to be perfected.
Elizabeth, I suppose a software error like the one that gave the Toyota accelerator trouble could happen with a self-driving car. That's possible, but in the meantime, about 30,000 drivers are dying on our highways every year. Cutting that down significantly would be worth the risk.
Rob, it's important to remember that Google's self-driving vehicles still require a driver to sit in the driver seat while the car is moving. Yes, the cars drive themselves, but it says something about the confidence level of the technology that drivers must still be ready to take over. Government bodies and consumers still don't have complete confidence in the technology.
Given that the folks paying for this study have a financial stake in the efforts to produce self driving vehicles, what else could any research group possibly say? The investors, and those invested in don't stand to profit if the prediction is made that it is a bad idea and an introduction into a major fiasco. And a lot of engineers see it as a really fun challenge, which it certainly would be. While it might possibly be achievable at some cost, given the degree of reliability of current automotive computerized systems, it will not be acieved at a reasonable price.
Of course, the economics of a vehicle system for driving military vehicles is a lot different, since most of the characteristics and situations for tactical military operations are quite different from the world that we live in.
Did we not just read on this site about a Tesla being destroyed by a piece of metal on the road surface? Well how is that driverless car going to react? How about the vehicle on either side or in back? How will these driverless vehicles interact with motorcycles? Even if the technology can solve some of those issues, why? I agree with Old Curmudgen, if you cannot be bothered with driving, take public transportation.
I work with numerous people who have a long commute and they all carpool, so three out of four days they do not have to drive. As Creedence Clearwater said back in the 60's: "rushing to the treadmill. rushing to get home, worry 'bout the time you save". I do not mind my commute unless I get stuck on a bridge with an accident up ahead, but big deal. There has to be many more vital things to spend money on.
Transfers the control of a large number of motion axes from one numerical control kernel to another within a CNC system, using multiple NCKs, and enables implement control schemes for virtually any type of machine tool.
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.