如今谷歌（Google）已经证明，自驾汽车可以行驶30万英里而不出现任何差错。说起来，这辆车在停车场确实曾发生过一起小摩擦，但当时是驾驶员在操作。这辆改装过的丰田普锐斯的车顶上集成了GPS、雷达和3D测绘相机技术，因而可以实时“看到”交通信号、道路情况和行人，令人印象深刻。从事这项研究的并不只有谷歌。2010年，四辆无人驾驶汽车参加了VisLab州际自驾汽车挑战赛，从意大利的帕尔马一路开到中国的上海世博会（the World Expo in Shanghai）会场，全程长达8,077英里。美国的加利福尼亚州上个月通过立法允许自驾汽车上路，成为第三个通过类似法律的州。（但目前还有个问题：法律要求必须有人坐在驾驶座上，随时准备在紧急关头接管汽车。）10月份，日产发布了聆风（Leaf）车型的自驾原型款，并表示该车将于2015年上市。通用汽车（GM）、福特汽车（Ford）、丰田（Toyota）和宝马公司（BMW）也在测试类似的车型。美国乔治华盛顿大学交通安全与防护计划（George Washington University's Transportation Safety and Security Program）负责人阿兹姆•埃斯坎达里安指出：“相关科技依然存在，但成本必须降下来。10到15年之后，人们会看到很多这种汽车。”

    不过，关于自驾汽车将如何影响商业世界，如何改变包括汽车厂商、汽车保险公司、运输公司和钢铁厂商在内的各行各业，以及在这一过程中谁将得益谁将受损，相关的分析还很少。这类汽车还可以促进燃油效率的提高。例如，GPS可以发现空余的停车位。据美国麻省理工学院媒体实验室（MIT Media Lab）发布的一份报告，“在拥堵严重的都市地区，约有40%的汽油浪费在了驾车族寻找车位的过程中。”

    专家认为，自驾汽车会提高生产力。根据毕马威（KPMG）和汽车研究中心（Center for Automotive Research）联合发布的一份关于自驾汽车的新报告，如今美国上班族平均每年在方向盘后面花去250个小时。如果所有这些时间可以用于回复电邮或（合法地）发送短信，而汽车可以自己驶向目的地，情况会怎么样？它在防范事故方面自驾车也有优势。想像一下吧，刚刚在公司假日派对上喝得东倒西歪的员工们被各自的汽车送回家。

    卡车运输行业也可能深深受益于这一科技。可以描绘这样一幅图景：排成长列或曰“车队”的自动驾驶的、18轮大卡车在州级高速公路上的特殊车道上以每小时100英里的速度呼啸而过，彼此之间的距离只有12英寸。汽车研究中心的专家理查德•华莱士估计，自驾卡车可以将燃油效率提高15%～20%。“没有驾驶员，意味着不用停车加油或补给食品，也不会有人整夜开着空调在车内睡觉。”

    在劳动力短缺的偏远地带，矿业公司可用庞大的卡车运输数以吨计的矿石，而无需操作员掌控方向盘。事实上，矿业巨头力拓（Rio Tinto）已经在澳大利亚西部部署了这种车辆。由于不再需要驾驶人员，每辆卡车每年可以节省多达10万美元的成本。

    现在的汽车已经包含约1亿条电脑代码，用以运行引擎、电子制动与转向系统以及相当于自驾汽车先驱的一些功能，比如车道偏航提醒系统和牵引力控制系统——可接管制动与转向以防打滑。随着汽车变得越来越自动化，软件需求也会随之增长，这对谷歌和英特尔（Intel，该芯片厂商正投入1亿美元，开发自驾汽车所用的电子设备）来说是好消息。

    Google has now proved that a self-driving car can travel more than 300,000 miles without a mishap. Well, it did suffer a parking-lot fender-bender -- but a human was at the wheel. Its customized Toyota Priuses use an impressive combination of GPS, radar, and a 3-D mapping camera on the car's roof that "sees" traffic signals, road lanes, and pedestrians in real time. Google (GOOG) is not alone. In 2010, as part of the VisLab Intercontinental Autonomous Challenge, four driverless electric vans made the 8,077-mile drive from Parma, Italy, to the World Expo in Shanghai. Last month California became the third state, after Nevada and Florida, to make self-driving vehicles street legal. (The catch, for now, is that a human must sit in the driver's seat, ready to take over in an emergency.) In October, Nissan revealed a self-driving prototype of its Leaf that the car maker says could hit the market by 2015. GM (GM), Ford (F), Toyota (TM), and BMW are experimenting with similar models. AzimEskandarian, the director of George Washington University's Transportation Safety and Security Program, says that "the technology is here, but the cost needs to come down. In 10 or 15 years you'll see a lot more of these cars."

    There has, however, been little thought about how autonomous vehicles might impact business and alter industries ranging from car makers to auto insurers to transport companies to steel fabricators, and in the process leave a trail of winners and losers. These vehicles could also help boost fuel efficiency. For instance, GPS could identify empty parking spaces. A report published by the MIT Media Lab states, "In congested urban areas, about 40% of total gasoline use is in cars looking for parking."

    Experts believe self-driving cars will increase productivity. According to a new report on self-driving cars by KPMG and the Center for Automotive Research, the average American commuter now spends 250 hours a year behind the wheel of a vehicle. What if those hours were spent answering e-mails or (legally) sending texts as the car drove itself? There are liability advantages too. Imagine tipsy employees driven home by their car from the office holiday party.

    The technology could be a boon to the trucking industry. Picture long lines, or "platoons," of self-driving 18-wheelers, 12 inches apart, speeding down a special lane on the Interstate at 100 mph. Richard Wallace, of the Center for Automotive Research, estimates that self-driving trucks would boost fuel efficiency by 15% to 20%. "No drivers, no stops for fuel and food, and no one sleeping overnight in the cab with the air conditioning running," he says.

    In remote areas, where labor is scarce, mining companies can use giant trucks carrying tons of ore without an operator at the seat. The mining giant Rio Tinto, in fact, has deployed exactly this kind of vehicle in Western Australia. Because it no longer needs drivers, it saves as much as $100,000 a year per truck.

    Cars already contain roughly 100 million lines of computer code to run engines, electric braking and steering systems, and features that are precursors of self-driving cars, such as lane warning systems and traction control, in which cars take over braking and steering to avoid skids. As cars become more autonomous, software needs will increase -- which is good news for Google and Intel (INTC) (the chipmaker is investing $100 million in electronics for autonomous vehicles).

    自驾汽车很少会发生碰撞事故（至少在理论上确实如此），因此车辆可以造得更轻型化，无需笨重的安全框架，这意味着对钢铁的需求减少。随着安全问题不再那么令人担忧，汽车厂商将来还可以自由地作出新尝试，重新考量“何为汽车”这个基本概念。“为什么不设计一辆相当于车轮上的办公室的汽车呢？”美国加州大学戴维斯分校交通研究中心（Institute of Transportation Studies at the University of California）创始人兼主任丹•斯珀林问道。

不出意料：自驾汽车的最大障碍是律师们。好消息是，这项技术应该会显著降低在全美各地公路上丧命的人数——目前这个数字每年超过3万人。根据美国汽车协会（AAA）的统计，美国人每年因车辆碰撞耗费高达2,990亿美元。可是，一旦自驾汽车真的出现事故，谁该负责任呢？消费者坐在方向盘后面，根本没动方向盘。相比他们，侵权律师们更可能盯上财力雄厚的汽车厂商和编写了相关代码的软件公司。斯珀林还指出：“自驾汽车带来了灾难性事故的风险”，这意味着汽车保险行业也得作出相应调整。随着事故发生率下降，且汽车厂商将会分担事故责任，车主们对汽车保险的需求将会下降。为了维持收入，汽车保险公司必须改变目标，转而向汽车厂商或编写相关代码的软件公司——取决于律师们更喜欢将谁作为被告——提供保险服务。

    自驾汽车的变革将是渐进的，但这一进程已然启动。我们已经看到了汽车中出现的小变化，比如车道偏航提醒系统和自适应巡航系统。或许在不那么遥远的将来，我可以在坐车上下班的途中写稿了。

    总而言之，自驾汽车从下面四个方面影响商业世界：

    1.卡车可以结成“车队”行使，彼此之间只维持12英寸的距离，从而降低风阻。燃油效率有望提高15%～20%.

    2.美国人平均每年在上下班路上花费250小时。如果汽车可以自动驾驶，这些时间以用来干活。

    3.由于自驾汽车将更为安全，它们将不再需要沉重的安全车厢，对钢铁行业来说这是个坏消息。

    4.保险公司得弄清楚谁来为交通事故负责：汽车厂商、软件设计公司还是GPS供应商？

    译者：黄兴宇

    Self-driving cars rarely crash (in theory, at least), so vehicles can be made lighter, without heavy safety frames, which means less demand for steel. As concerns over safety diminish, car makers will have free rein to redesign the very notion of what a car is. "Why not design a car as an office on wheels?" asks Dan Sperling, the founding director of the Institute of Transportation Studies at the University of California at Davis.

    Perhaps the biggest obstacle facing self-driving cars is, not surprisingly, lawyers. The good news is that this technology should dramatically reduce the 30,000-plus annual fatalities on the nation's highways. According to AAA, traffic crashes cost Americans $299 billion each year. But who will be responsible when accidents do happen? The deep-pocketed automakers and the software company that wrote the code are more likely targets for tort lawyers than the consumer sitting behind an (unused) wheel. Sperling says that "self-driving cars create the risk of catastrophic accident," which means the auto-insurance business will need to reboot. As accident rates fall and liability becomes shared with the manufacturer, drivers will see less need to carry so much insurance. To maintain their revenue stream, auto insurers will have to shift to providing coverage to the car makers or software firms that write the code -- depending on which tack the trial lawyers take.

    The change to self-driving cars will be evolutionary, but the evolution is well under way. We're already seeing small changes in our vehicles, such as lane warning systems and adaptive cruise control. Maybe in the not-too-distant future, I'll be able to write stories like this while commuting in my car.

    Four ways self-driving vehicles may change business

    1. Trucks could travel 12 inches apart in "platoons" that reduce drag. Fuel savings could reach 15% to 20%.

    2. Americans spend on average 250 hours a year commuting. If a car does the driving, that time could be spent working.

    3. Because self-driving cars will be safer, they won't need heavy safety cages. Bad news for the steel industry.

    4. Insurers will have to figure out who's liable in an accident: the car maker, the software designer, or the GPS provider.