这是Google的自动驾驶汽车，真正的概率机器人应用。
Sebastian Thrun：
Sebastian Thrun is a Research Professor of Computer Science at Stanford University, a Google Fellow, a member of the National Academy of Engineering and the German Academy of Sciences. Thrun is best known for his research in robotics and machine learning, specifically his work with self-driving cars.
Thrun 还写了一本教材《Probabilistic Robotics》，是概率机器人领域的经典。
Thrun's TEDTalk：
Text:
All it took was a roughly 60-second lap around the top of a parking structure at high speeds to convince TED attendees that autonomous vehicle technology has come a long way. Navigating hairpin turns with precision unmatched by many human drivers, the car in Long Beach merely hinted at its real potential. I'm excited to say that progress on the project has continued rapidly since I was at TED, and I'm more motivated than ever to realize what has been my dream and that of many others for a long time.
Like most people, my connection to driving is very personal. I lost my close friend Harald to a car accident when I was 18, after he lost control on an icy road and collided with an oncoming truck. Then, roughly a year ago, my lab manager Suvan was killed after a distracted driver hit her Prius from the side while she was crossing an intersection. Sadly, these tragedies are not uncommon.
In fact, the statistics are sobering. After the U.S. Department of Transportation recorded 32,788 traffic fatalities in 2010, Secretary of Transportation Ray LaHood referred to distracted driving as "a deadly epidemic." Think about all the distractions you face when driving -- the kids in the backseat, a ringing cell phone, the stalled vehicle on the side of the road. These moments are part of life and therefore are hard problems to solve. When you have a very hard problem, the key is sometimes to approach it from an unexpected angle. In the first legislation of its kind from any state or nation, the state of Nevada recently passed a law paving the way for drivers to operate a self-driving, autonomous car.
As you might expect, this means that a vehicle may drive itself without the active control of a human operator. More specifically, the car technology we have been working on at Google uses radar, GPS, cameras, a rapidly spinning LIDAR laser, and other devices to navigate properly and understand a complex driving environment.
I believe wholeheartedly that self-driving cars can make traffic safer. After all, they never blink or focus on switching the radio station. They have no blind spots, and they can simultaneously process information at a rate that no human can match. Driving is a completely different experience when you know the exact speeds and locations of all the vehicles around you, and when you can interpret and call upon detailed maps of road infrastructure and street-level signs. These capabilities may one day open up driving to the elderly and disabled who today do not have driving privileges. It's an exciting future.
But self-driving cars are not just about safety. I estimate that the average car is immobile 97% of its lifetime. Think about how many occupied parking spaces and underused cars that means. With less than 5% of Americans using public transportation to get to or from work, are we tackling this problem correctly? Self-driving cars can open up new models of car sharing that we can only imagine today. One way or another, America should take the helm of this kind of innovation.
Most of my professional life has involved my attempts to turn science fiction into reality. Thanks to the hard work of a world-class team, our self-driving cars have completed nearly 200,000 miles of autonomous driving on public highways. I envision a future in which this technology is available to anyone, and that's a vision I would like us all to keep on our horizon.