無人駕駛汽車的未來不再遙遠

JERUSALEM — Last month, on a freeway from Jerusalem to the Dead Sea, I sat in the driver's seat of an Audi A7 while software connected to a video camera on the windshield drove the car at speeds up to 65 miles an hour — making a singular statement about the rapid progress in the development of self-driving cars.

耶路撒冷——上月，在從耶路撒冷到死海的高速路上，我坐在一輛奧迪(Audi)A7的駕駛座上，而與裝在擋風玻璃上的攝像頭相連的軟件，則在駕駛這輛汽車以每小時65英里（約合104.6公里）的速度前進，顯示着自動駕駛汽車發展的神速。

While the widely publicized Google car and other autonomous vehicles are festooned with cameras, radar and the laser range finders called lidars, this one is distinctive because of the simplicity and the relatively low cost of its system — just a few hundred dollars' worth of materials. "The idea is to get the best out of camera-only autonomous driving," said Gaby Hayon, senior vice president for research and development at Mobileye Vision Technologies, the Israeli company that created the system in the Audi.

受到廣泛宣傳的谷歌(Google)汽車和其他自動駕駛汽車裝滿了攝像頭、雷達和被稱作“激光雷達”(lidar)的激光測距儀，而這部汽車的獨到之處卻在於簡潔，以及成本相對較低的系統，材料只需幾百美元。爲奧迪研製這套系統的以色列移動眼視覺技術公司(Mobileye Vision Technologies)負責研發的高級副總裁加比·阿永 (Gaby Hayon)稱，“我們的想法是充分開發只用攝像頭的自動駕駛系統。”

The Mobileye car does not offer the autonomy achieved by Google's engineers. The Google car, which has been tested for more than 300,000 miles in California traffic, will merge onto freeways, drive safely through intersections, make left and right turns, and pass slower vehicles.

使用移動眼的汽車不具備谷歌工程師所實現的那種自動化程度。谷歌汽車曾在加利福尼亞進行過30萬英里的駕駛測試，它將會出現在高速公路上，在交叉路口安全行駛，能左右轉彎，還能超過速度稍慢的汽車。

By contrast, the Mobileye vehicle is capable only of driving in a single lane at freeway speeds, as well as identifying traffic lights and automatically slowing, stopping and then returning to highway speeds.

相反，移動眼汽車只能以高速在一條車道上行駛，可以識別紅綠燈，並自動減速、停車，之後再繼續以高速行駛。

But by blending advanced computer vision techniques with low-cost video cameras, the company is demonstrating how quickly autonomous driving can be commercialized. "You cannot have a car with $70,000 of equipment," said Amnon Shashua, a computer scientist at Hebrew University and a founder of Mobileye, referring to Google's lidar system, "and imagine that it will go into mass production."

然而，通過將先進的計算機視覺技術與低成本的攝像頭搭配使用，這家公司向人們展示了，自動駕駛的商業化可以多麼快地實現。希伯來大學(Hebrew University)計算機科學家，移動眼公司創始人阿姆農·沙舒瓦(Amnon Shashua)評價谷歌的激光雷達系統說，“要是一輛汽車搭載的設備價值高達7萬美元，你是沒辦法設想讓它大規模生產的。”

Elon Musk, the automobile and spaceship entrepreneur, recently made headlines by saying much the same thing. Mr. Musk said he was interested in designing self-driving versions of his Tesla automobiles, possibly in partnership with Google, but in an interview with Bloomberg Mr. Musk said lidars were too expensive to be used in production cars. Like Google, Mobileye is not an automobile company, but rather a technology firm that is intensely focused on developing the next generation of artificial intelligence software.

汽車和宇宙飛船行業的企業家埃隆·馬斯克(Elon Musk)最近因發表大致相同的看法而上了新聞。馬斯克稱，他很有興趣爲爲自己的Tesla汽車設計自動駕駛的版本，可能會與谷歌合作，但在接受彭博社(Bloomberg)採訪時，他表示激光雷達成本太高，不適用於汽車製造。與谷歌一樣，移動眼也不是一家汽車公司，而是主要研發下一代人工智能軟件的技術公司。

The company was founded in the 1990s, after developers persuaded General Motors to buy an inexpensive camera that could detect vehicles in adjacent lanes.

20世紀90年代，開發人員在說服通用汽車(General Motors)購買一種可以探測到相鄰車道上的汽車的廉價攝像頭之後，成立了這家公司。

Mobileye has recently begun offering the third generation of its technology, which companies like Volvo have promoted for its ability to detect pedestrians and cyclists. Nissan also recently gave a hint of things to come with a demonstration of a car that could automatically swerve to avoid a pedestrian who walked out from behind a parked car. The system was based on Mobileye technology.

最近移動眼公司開始提供第三代技術，由於能夠探測行人和騎自行車的人，這項技術得到了像沃爾沃(Volvo)這樣的公司的推廣。日產(Nissan)最近展示的一款汽車，可以自動轉彎避開從停着的汽車背後走出的行人，其系統的基礎就是移動眼技術。這款車可能暗示了未來的某種方向。

As soon as this summer, the first limited systems offering a feature known as "traffic jam assist" will begin arriving from more than five major automobile makers. Those cars will drive safely in stop-and-go traffic, but will require that drivers keep their hands on the steering wheel.

今年夏天，首批具備有限功能，提供“堵車助手”服務的系統就會裝配到五家以上大型汽車製造商生產的汽車上。這些車可以在走走停停的路況中安全行駛，但司機還是需要把手一直放在方向盤上。

But more advanced systems will be introduced as early as 2016, according to Mobileye, and it was that advanced capability I experienced last month, during my foray in freeway traffic.

不過，移動眼透露，更先進的系統最早將於2016年推出，它正是我上個月在高速路上飆車時，體驗的那套功能先進的系統。

With two Mobileye engineers, Dr. Hayon and Eyal Bagon, we drove several miles east of Jerusalem, where Mobileye is located, and then pulled off at a nondescript turnout where another Mobileye employee waited in a shiny white Audi A7.

在兩名移動眼工程師阿永博士和埃亞勒·巴貢(Eyal Bagon)博士陪同下，我們在耶路撒冷以東開了幾英里路，那裏是移動眼公司的所在地，之後在一個沒什麼特徵的岔道停了車，另一名移動眼的員工坐在一輛閃閃發亮的白色奧迪A7裏等我們。

On the way out of town, the two men were apologetic. This was a work in progress, they told me.

在出鎮的路上，兩名男士充滿歉意，他們對我說，這是一套還在研發的系統。

I was already feeling a little let down. In California in 2010, I had been the first reporter to drive in the Google car, a Toyota Prius fitted with sensors that created a remarkably detailed map of the world around the car. Moreover, it was connected wirelessly to the Internet, giving it access to a vast cloud-based set of data that could be matched to what the local sensors were seeing.

我已經感到有點兒失望了。我曾是2010年在加利福尼亞駕駛谷歌自動駕駛汽車的第一個記者，那是一輛豐田普銳斯(Toyota Prius)，上面佈滿感應器，它們針對車周圍的世界，創造出一幅細緻得難以置信的地圖。不僅如此，它還能無線上網，這讓它能獲得以雲計算爲基礎的大量數據，這些數據能夠與現場感應器正在捕捉的圖像一一契合。

It was a tour de force. The car merged seamlessly with traffic on Highway 101 and then turned off the freeway to automatically drive through downtown Mountain View, Calif., where Google has its headquarters. It was taken out of automatic pilot just twice — while passing a cyclist and then again to back into a parking space. The Google engineers did not want to chance a mishap with a reporter on board.

那是一件傑作。車天衣無縫地融入了101號高速路的車流，而後從高速上下來，自動駕駛着穿過加州山景市區，即谷歌公司總部所在地。車子退出自動駕駛模式的情況只發生了兩次，一次是繞過一個騎自行車的人，另一次是倒車進入停車位。鑑於有一名記者在車上，谷歌工程師不想冒發生事故的風險。

There was no such caution last month with the Mobileye engineers.

上個月，移動眼公司的工程師就沒這麼小心翼翼。

"Why don't you drive?" Dr. Bagon suggested as he slid into the passenger seat behind a large display and keyboard. Dr. Hayon jumped into the back seat, and I had no choice but to get behind the wheel.

巴貢坐進前面裝着大顯示器和鍵盤的副駕駛時建議，“要不你來開？”阿永已經跳進了後座，我別無選擇，只能坐進駕駛座。

The Mobileye engineers proceeded to give me a rapid-fire lesson on how to drive their robot car: Turn on the automatic navigation by pressing the cruise-control button and then pulling a stick on the steering wheel toward you to activate the "lane-keeping," or steering, feature.

移動眼公司的工程師隨即給我上速成課，教我如何駕駛他們的自動車：按下巡航控制按鈕，打開自動導航，然後把方向盤上的操縱桿扳向自己，啓動“車道保持”或駕駛模式。

Following another engineer who had taken off in a car ahead of us, we pulled onto the highway. After a false start, I engaged the autopilot, and off we went; I put my hands in my lap and tried to relax as we snaked down a desolate canyon road on our way to the Dead Sea.

另一名工程師坐在我們前面的一輛車裏，我們跟着他駛上了高速公路。第一次啓動失敗後，我打開了自動駕駛儀，然後汽車便開動了；我把雙手放在大腿上，試圖在沿着一條荒涼的峽谷道路前往死海的途中放鬆片刻。

In the Google car, I had felt detached from the machine intelligence that guided it. But in the Mobileye car I was acutely aware of the autopilot's limitations. The car had a tendency to weave a bit when it started to pull away from an intersection. Once, as we were passing a parked car, the Audi pulled in the direction of the other vehicle. I lifted my hands out of my lap and nudged it back to the center of the lane. The Mobileye engineers sat perfectly calm.

坐在谷歌的汽車內，我覺得自己與引導車輛的機器智能並沒有太大關係。而坐在移動眼車內，我能夠明顯覺察到自動駕駛儀的侷限性。在離開交叉路口的時候，行車的路線總會有點迂迴。有一次，當我們開車經過一輛停靠在路邊的汽車時，奧迪車朝另一輛車的方向駛了過去。我擡起放在腿上的手，把車重新開回路中央。移動眼公司的工程師非常冷靜地坐在一旁。

After 20 minutes of freeway driving, I pulled the Audi off at an intersection, and then put it back on autopilot and headed toward Jerusalem. The demonstration was not as eye-popping as my Google ride, but it gave me a clearer understanding of what the automobile industry has in its sights.

在高速公路上行駛了20分鐘以後，我把奧迪車停在了一個岔路口，然後打開自動駕駛儀，朝耶路撒冷駛去。這次試駕並不像駕駛谷歌汽車那樣令人震驚，但它讓我更加清楚地理解了汽車行業的願景。