[Netease Smart News, April 26th] The game about self-driving cars started. Google, Tesla, and most automakers such as Toyota, Volvo and Ford have thrown billions of dollars in hopes of getting their own self-driving cars on the road as quickly as possible. Earlier this month, Apple became the 30th company approved to test driverless cars on public roads in California. According to reports, even Amazon has created a team to understand how to better use this upcoming transportation revolution. Waymo has already begun testing its self-driving car in Phoenix. But the road to full autonomous driving is long, tortuous, and dangerous. In December last year, Uber’s self-driving car was found to have a red light after driving for a few hours in San Francisco. In Florida last year, a Tesla driver died in a car accident when the car was using an autopilot system and crashed into a semi-trailer. Christopher Hart, former chairman of the National Transportation Safety Board of the United States, recently said: "I don't think that the full automation of cars will soon be realized." He said: "No software designer in the world can foresee all possible situations... Dogs on the street, people on the street, cyclists, police or construction workers." So, in the coming years, will we only use artificial driving? I don't think so. However, if we want to take our hands off the steering wheel in a short time, we should stop focusing our attention on the car. Instead, we should build smart lanes dedicated to self-driving cars on existing highways to isolate any uncertain behavior. Here's how it works. First, we should correct the highway that has been equipped with smart sensors and road signs. Then, we should use barriers to completely isolate a lane so that it is only suitable for vehicles that have the necessary technology (such as 360-degree sensors, radar, car-vehicle radios, automatic steering, and adaptive cruise control). Barriers and other things pass. When the vehicle enters and exits the lane at the specified location, the system assigns a position to each vehicle in the traffic stream. The entire traffic will be cruising at high speeds, and each car will be followed by a car. With onboard wireless devices, each car and highway sensor will become part of the same mobile network, such as bypassing obstacles or detecting the presence of interference in advance. In the case of a team, the car can save fuel (only the head car will encounter headwinds), and the driver - or more precisely, the passenger - can sit and relax. Technology can already be achieved. The agreement on car-to-car wireless communication (V2V) and vehicle-infrastructure wireless communication (V2I) will soon become a necessary configuration for all new cars, and a technology wave similar to wireless WiFi will soon embed traffic lights and parking signs. And work area, allowing two-way communication between the car and the infrastructure. The idea of ​​building an autonomous driving fleet is not new: Swedish truck manufacturer Scania recently announced that it will deploy a fleet of self-driving trucks on one of Singapore's roads, while Mercedes has tested a three- truck fleet from Stuttgart to Rotterdam. A total of 375 miles. Technology companies often don't have the patience to work with regulators: They like to develop, transfer, test, and iterate, and change the current legal boundaries where they are needed, just as Uber is currently doing. However, technological slogans such as “quick action and breaking existing things†are more effective in social networking than in the real world, because in the real world, humans themselves are easily broken and easily fragile. Intelligence is often proportional to the scale of the application, and the urban road contains infinite variables. They are best handled by a highly adaptive (even easily distracting) human brain, rather than a machine that can only handle a predetermined scene. In the coming years, today's smart cars are still unable to cope with urban roads and unpredictable human behavior. But starting today, they can safely travel hundreds of miles on the smart highway, creating an important springboard for a wider range of applications. How much does this cost? We can refer to the toll highways: depending on the complexity and mileage, new toll highways may cost anywhere from a few hundred million to several billion US dollars, usually through public-private partnerships. This may seem high - but if we see the impact of automatic cars on the U.S. economy, we will think that it is not much. Engineers at the University of Texas estimate that this impact on the overall U.S. economy will reach 1.2 trillion U.S. dollars. When a company like Intel spent $15.3 billion to acquire the leading position of the connected car industry in an acquisition, it is difficult to understand why the U.S. federal and local governments cannot use a small part of it to help build future transportation. If President Trump can fulfill his pledge to improve the U.S. infrastructure funding by 1 trillion U.S. dollars, it will have a budget to build a smart highway. Internet car market (English source /venturebeat compiler / machine Xiaoyi proofread / 晗 ice) 燑br>