When the Defense Advanced Research Projects Agency's (DARPA) Grand Challenge autonomous vehicle race debuted in 2004, the results bordered on comedy. None of the vehicles drove more than 7.4 miles. Some ran in dizzy circles. Others stalled atop rocks. A robotic motorcycle journeyed all of two feet before toppling over. A 10-ton commercial truck lumbered forward a few yards and stopped like a confused buffalo, overwhelmed by the multitude of sensor signals zipping past its small brain.
Their technical struggles didn't last long. Nineteen months later, five vehicles finished the 132-mile second Grand Challenge. In 2007, six more completed DARPA's 60-mile Urban Grand Challenge.
Today the work continues. Google's automated cars have logged more than 140,000 miles. General Motors has predicted that self-driving vehicles will be ready by the end of this decade.
Click the image below to see our slideshow of the evolution of autonomous vehicles over the past seven years.
Stanford University's "Stanley" won the 2005 Grand Challenge by finishing in six hours, 53 minutes. The Volkswagen Touareg R5 combined six laser range finders with a color camera, radar system, GPS system, and inertial navigation system.
Photo courtesy of Stanford University
For Further Reading
To keep up with our Chevy Volt coverage, go to Drive for Innovation and follow the cross-country journey of EE Life editorial director Brian Fuller. On his trip, sponsored by Avnet Express, Fuller is driving a Volt across America to interview engineers.
While it's one thing to see these crazy vehicles as part of DARPA development projects, it's quite another to think of an autonomous car like the one Google is working on fighting New City cab drivers for right of way, crawling through the traffic jams in San Francisco, or cruising (hopefully not careening off) over the Golden Gate bridge.
As one that doesn't even trust the rear-view camera for backing out or parallel parking, I'm not sure I could leave the driving to the car--no matter how much onboard intelligence you pack on.
I'm a bit confused by the further reading links provided. While all-electric cars are admirable, I'm not quite sure how they are related to self-driving cars. Stanley is a diesel-powered Volkswagen, while the Google self-driving cars are based on the hybrid Toyota Prius, the same platform used for their fleet of Google Street-view cars. I'm all for innovation and I am hesitant to disparage any engineering lab, but I find it difficult to believe that the self-driving car revolution will be led by General Motors. I've read GM's press releases but I haven't seen GM take the technology lead in the past. Please set me straight if I'm way off base.
That is right, I won't be able too be fully confident in letting my car drive on itself. Though the texting while driving is funny yet dangerous, that's the one causes major fatal accidents on the road. Yeah, I agree that General Motors is not yet on the verge of technological advancement to lead this kind of automobile innovation.
I have spent a few hundred hours at the testing end of automotive production lines, as well as watching the electrical portions of the vehicles being built. My experience says that no way would the quality be high enough for a life-critical system. We could have driver warning systems that could enhance safety, but production quality would just not be adequate for a road-speed autopilot.
Besides that, the programs would probably be written by some of the same individuals that write the current vehicle system code. That means that the autonomous vehicle would not think at all in the manner that I think, and so it would undoubtedly be unable to handle any exceptions to what it expected. So leave the computers to the engine controls and the navigation screens, where the harm that they can do is much less.
I agree it is definitely a fun challenge, but I do not know how much confidence I would have in a vehicle driving itself. Cars are already on the market that park themselves, but i am not sure if they will be able to market a car that drives itself by the end of this decade.
GM says they will be ready by the end of the decade. I can believe it. There are already mining trucks, ag tractors and other equipment that operate autonomously. Highway speeds with other drivers are a different environment since there is much more variability. these vehicles need to be capable of reacting to random human behaviour, just like we are when at the wheel.
We already have cars on the road, and every time one of these vehicles encounters a situation it isn't programmed for, we can add that to the abilities of the control system. The testing required to deploy these vehicles in production will be in the tens of millions of miles I would guess. A daunting task, but acheivable.
In the early days, there may be vehicles with limited situational capability. Lets say we have a vehicle that can handle highway traffic, but will require the driver to take over in any sort of urban environment with stoplights and such. This capability still has value without being absolutely autonomous.
Wow – we already have Skype (so glad there is a "no video" option - I think Jane would have preferred that for those early morning calls) and now yet another futuristic aspect of The Jetsons working towards reality. Now if I could only get a household robot like Rosey, my German Shepherd could learn to speak like Astro and we could all go to a three hour/three day work week like George – now that would be progress!
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.