Xinhua News Agency report of November 22 from Shenzhen: Artificial Intelligence Changes Traditional Driving Intelligence How Far Is Traffic from Us?
Xinhua News Agency reporters Chen Yuxuan, Wang Xiaodan and Lv Qiuping
Fatigue driving will trigger an early warning system, set the route for cars to be driverless, and traffic lights can be adjusted according to traffic flow. Artificial intelligence is changing traditional driving behavior. The 19th China International Hi-tech Fair ended in Shenzhen on the 21st. At this fair, the deep integration of artificial intelligence and urban transportation has attracted the attention of visitors: how far is intelligent transportation from us with unmanned driving?
Artificial Intelligence Changes Traditional Driving Behavior
You may have seen a scene where the driver yawned one after another, dozed off like a chicken pecking rice, and his eyelids grew heavier and heavier until he could hardly open them. Fatigue driving is one of the important causes of traffic accidents in major cities. Now, the application of artificial intelligence in the field of driving may reduce the threat of driver fatigue to road traffic safety.
At this high-tech fair, Shang Tang technology, which is engaged in the research and development of computer vision technology, demonstrated the latest driving monitoring and early warning system: in the future, cameras embedded with face recognition algorithms can read and analyze drivers' facial expressions in real time. once the driver's fatigue degree exceeds a reasonable range, the system will send out early warning signals to remind drivers to pay attention to safety.
Not far from the Shangtang technology booth, a car loaded with driverless systems attracted many visitors. In the first half of 2017, Tong Xianqiao resigned from the United States and returned to Shenzhen to start his own business, pushing the driverless system jointly developed by himself and two friends to the domestic market. Tong Xianqiao said that their unmanned system has been tested on the road using domestic sensors and self-developed core algorithms. "In the Silicon Valley road test before, our car drove on the highway with normal traffic flow for about 30 minutes and all the way was safe." He said.
At the 2017 Baidu World Congress held recently, Baidu Chairman Robin Li said that it is expected that the unmanned minibus cars jointly produced by Baidu and Jinlong Bus will be mass produced in July 2018. In addition, Baidu will also launch more driverless cars with JAC, BAIC and other auto manufacturers.
"City Brain" Takes First Step in Intelligent Transportation
Artificial intelligence is changing traditional driving behavior and will also bring a complete set of urban intelligent transportation systems. In Hangzhou, artificial intelligence has been applied to the overall traffic control of the city.
In October 2016, ariyun company launched the "city brain" pilot project. The pilot uses the Internet as the infrastructure and makes full use of the abundant urban data resources to carry out overall real-time analysis of the city, which can effectively allocate public resources and promote the sustainable development of the city.
At the forum held at this year's hi-tech fair, Wang Jian, chairman of Alibaba group's technical Committee, introduced the operation of the "city brain": in Hangzhou, the "city brain" uses cameras to analyze real-time traffic flow, allowing traffic lights to adjust according to the real-time traffic flow, optimizing the time allocation at intersections, and improving traffic efficiency.
According to the data, after Hangzhou implemented the "city brain", the average driving speed in the peak period in the pilot areas increased by 15%, and the average congestion time in the areas decreased by 9.2%. Xiaoshan District of Hangzhou City has also implemented the priority dispatch of 120 ambulances and other special vehicles. Through the "lifeline" of green lights customized by the "city brain", the time for ambulances to arrive at the scene has been shortened by nearly half.
For congestion, illegal parking and accidents, the "city brain" can also replace artificial real-time detection and trigger mechanism for intelligent processing. In the main urban area of Hangzhou, the average daily alarm number of "urban brain" events reached more than 500, with an accuracy rate of 92%.
Peng progress, chief architect of yushi technology, who is engaged in the research and development of unmanned technologies, said: "artificial intelligence is bringing urban transportation into a new era of higher efficiency and safety."
Prospects Worth Expecting and Difficulties to Be Breakthrough
At present, "artificial intelligence+transportation" has become a hot spot in the capital market. The report from Tencent Research Institute shows that in China's artificial intelligence enterprises, the total amount of financing in the fields of auxiliary driving and automatic driving ranks third, reaching 10.7 billion yuan, second only to the fields of computer vision image and natural speech processing. According to the analysis of the report, since there are only about 31 enterprises in the field of auxiliary driving and automatic driving in our country, but they have the third largest amount of financing, which means that investors are very optimistic about this field.
Investors' confidence comes from China's unique advantages. Peng said that China's major cities have rich traffic scenes and huge amounts of traffic data, which provide unique conditions for the training of unmanned driving algorithms.
However, it is still difficult to realize the deep integration of artificial intelligence and traffic. Peng said that unmanned driving is the core scenario of artificial intelligence application in the transportation field. It is actually a comprehensive problem involving the overall transportation operation of the city, including environmental perception, intelligent decision-making and planning, intelligent control and other fields. It will take some time to make breakthroughs in these related fields.
JASON ZHANG, product manager of Shenzhen Solling Industrial Co., Ltd.' s Vehicle Networking Business Center, said that in order to realize intelligent transportation, in addition to unmanned driving, it is also necessary to solve the communication problems between vehicles and vehicles, and between vehicles and clouds.
"In the future, we will use sensors, high-definition cameras, lidar and other tools to collect real-time information of road infrastructure, traffic lights and vehicles, and the algorithm will make correct instructions based on the information." He said.
A new round of exploration has begun. Recently, the Ministry of Science and Technology decided to rely on Baidu to build a new generation of autonomous national artificial intelligence open innovation platform, while the "city brain" national new generation of artificial intelligence open innovation platform will be undertaken by Aliyun Company.