CHINESE ENGLISH
Industry News

Location:Home > News > Industry News... > Artificial intellige...

Artificial intelligence travel is easier.

Source:Updated:2018-03-06 08:16:47

From driverless cars to online car-hailing, from path planning to traffic management, change is everywhere.

You don't have to keep your eyes on the front, your hands don't need to grip the steering wheel, your feet don't have to step on the gas pedal, and the driver can do a "hands-off" when the vehicle is moving... Such "good things" are speeding up as driverless technology matures.

"At the moment, our smart cars have been able to drive autonomously in special sections such as low-speed congestion and expressways, and we can achieve mass production next year. After 7-10 years, it can be applied to at least 80% to 90% of the road conditions. Shen Haiyin singularity automobile founder and chief executive, said the unmanned in plain English is to make the car environment perception, path planning and its own autonomous vehicle control technology, and the key is behind this by artificial intelligence, deep learning "teaching" with various types of sensor vehicles to learn to drive.

From driverless cars to online car-hailing, from path planning to traffic management, the huge shift in artificial intelligence to transport is now ubiquitous.

"Open the APP and call the car, and the system will be able to predict the destination based on the user's historical record, and it will be able to pop up the recommendation." Drops travel, chief technology officer aoj out his phone, told reporters before the passengers before the general network such as car driver hit two phone to determine on position, while artificial intelligence history on site, tracking trajectory, according to the experience to determine early on site. At present, didi has excavated 30 million such sites in more than 400 cities.

Smart pie, one of the technology cores of the didi platform, is also seeing significant changes in artificial intelligence. A few years ago by passengers drops around the car, the system sends a mainly consider location, but the location is not equal to pick the shortest driving time recently, and then send list and added path planning, estimated time of arrival, owners service grading factors, multidimensional calculate the matching degree of passengers and the owner.

By now, didi has been more intelligent: taking into account the supply and demand of the platform and the actual situation of the road, matching orders and vehicles every two seconds to make a global optimal match.

"Artificial intelligence enables path planning to achieve a transition from rule-based to data-based." Scott map, vice President and chief human officer zhen-ning dong told reporters, path planning is one of the hardest road weight calculation, 2013 years ago, they were mainly, according to judge by the road distance and other physical conditions and after the introduction of artificial intelligence, can through to deep learning of users to track the huge amounts of data, iteration, so that the path planning more efficient, time prediction is more accurate.

Using the same principle, gaud's prediction accuracy for the estimated arrival time has also improved greatly in recent years, reaching 80% to 90%.

At the same time, companies such as gd and didi are also actively working with wuhan, jinan and hangzhou to optimize the traffic lights in some parts of the city. Previously, the intersection traffic light interval is usually fixed length, but by artificial intelligence have real-time perception, the traffic time interval can be dynamically adjusted, makes the downstream upstream team go, just the green light to release to the intersection.

Didi, for example, has optimized more than 800 lights in the country as of December 10, 2017, saving 10 to 20 percent less time than before.

Travel is more green, and the efficiency and safety of transportation resources are greatly improved.

"What is congestion? It is the uneven use of road resources. Solve the congestion, on people's experience, to lay down more roads alone is useless, the best way is through the artificial intelligence scheduling, realization of people, vehicles, road together, improve the efficiency of using the road resource. "Said dong.

Artificial intelligence makes travel more efficient and makes it more relaxed and green.

 

"Good self-driving technology will reduce fatigue, improve driving experience, and improve safety." "Said shen. Some experts say that people who drive cars need a total of 1 when they see an emergency and put the brakes on to make the brakes work. Two seconds, and the whole process of the driverless car is going to be zero. 1 second to 0. For 6 seconds, emergency braking is faster and driving is safer.

"When the Shared travel network is sufficiently developed, it is possible to use fewer cars to meet the needs of more people and greatly reduce the consumption of energy and resources." Aoj said that drops the greatest wealth is big data, artificial intelligence, machine learning algorithm, "we hope the real-time traffic information integration, by complex algorithm and computing platform for real time, intelligent decision."

"Advanced sensing technology, position and geographic information technology and wireless communication technology in the traffic application, the depth of the rapid increase of computer storage and processing ability of infrastructure, such as vehicle information level enhances unceasingly, for the application of artificial intelligence in the field of transportation laid the foundation conditions." Li bin, deputy director of the highway science research institute of the ministry of transport, concluded that the changes were mainly reflected in three aspects:

One is the amount of traffic data exponentially explosive growth, based on the depth of the traffic data of machine learning application provides a possible, commonly known as "traffic mind", realizing the maximization of transportation resources use efficiency.

The second is the intelligent management, organization, dispatching and control system of traffic flow, passenger flow, freight flow and passenger and freight transportation, which can make the most convenient travel and reduce the logistics transportation cost.

The third is the unmanned system in the complex environment, which will break through the limitation of human perception and reaction capacity, and greatly improve the transportation safety level and transportation efficiency.

Massive data and application scenarios, "AI+ traffic" in the future is infinite possibilities.

Development of "AI+ traffic", where is China's biggest advantage? In the respondents' answers, there are two things that are very consistent, one is massive data and the other is a huge application scenario.

The amount of data is unquestionable. At present, gaode has 700 million users, and the location service has been adopted by more than 300,000 apps, with more than 1.1 billion mobile terminals covered each month. Every day, the original data of didi's new track is over 70TB, and the daily processing data is 4500TB, and the daily route planning request is over 20 billion times.

More gratifying is that many enterprises and local governments have launched deep cooperation to gradually open data bottlenecks and "information islands" to dig deeper into data value. Beginning in July last year, Scott traffic brain maps and hangzhou cooperation projects, using big data of gold, can quickly detect urban congestion location accurately, and push it to the traffic police department, to speed up the police. According to the data, the accident can be reduced by 20% and the average speed by 35%. At the same time, the government will share information such as traffic control with gode to prompt users to detour.

"There are 20 million people in Beijing, 5 million cars, the roads are already saturated, and the need for congestion and Shared travel is more urgent." Jaco said americans for travel quality expectations will generate more traffic project, will also be reversed transmission with artificial intelligence to speed up the application in the field of traffic, "under the drive of application scenarios, any 1% efficiency will bring more influence to improve."

With the advantages of talent reserve and technical force, China's transportation enterprises are full of confidence in the future. In zhang's view, "over the past 10 years, we have solved the problem of information flow, and the future focus is on optimizing the flow of people and objects in the physical world."

"Looking into the future, ai will have at least three breakthroughs in the depth of transportation systems." The first is to break through the limits of human perception and reaction capacity, to realize the intelligence of vehicles, and to reduce the number of accidents and casualty rates. The second is to break the restriction of traffic flow theory, realize traffic automation through generalized control, and multiply the carrying capacity of road network. The third is to break the constraint of information asymmetry and realize the efficient utilization of transportation resources through the realization of the optimal transformation from individual and local optimal to the global optimum through resource allocation.

In order to achieve these visions, in addition to technological breakthroughs, we need to cooperate with the optimization of the facility environment. "Transport infrastructure in the future will be more information, along the road to deploy wireless communication and sensor system, and also want to better adapt to the self-driving car and someone is driving a car state of mixed traffic flow." "Li bin said.

Recently, the ministry of transport and clear, to grasp the new needs, such as automatic driving to promote digital, snatched, intelligent transportation infrastructure development, speed up the construction and the forming equipment and collaborative digital transportation infrastructure facilities, and organize to carry out autonomous, unmanned ship, solar road surface, such as research in cutting-edge technologies and tracking, study and formulate policies to promote the development of self-driving file, promote auxiliary self-driving technology in the field of operation of the vehicle to demonstrate the application.

"The pace of innovation in the era of artificial intelligence is completely different from that of the industrial age. As long as cutting-edge technology breaks through, the next step will be rapid." "Shen said with confidence.

Product|Technical|News|Propaganda|Partner|About

Copyright © 2014 Quanstar Intelligent Controls(shanghai)Co.LTD All rights reserved ICP: Shanghai ICP No. 10214886

Address: 152 Ring Road, Shanghai Comprehensive Industrial Development Zone

Tel: 021-57472600    E-mail:Mark.liang@quanstar.cn

Website buildingTrueland