January 21, 2022

Driverless technology gradually matures AI map traffic is more convenient

From drones to network rides, from path planning to traffic management, change is everywhere

Eyes do not have to keep an eye on the front, hands do not need to hold the steering wheel, and the feet do not have to step on the accelerator. When the vehicle is in motion, the driver can completely be a “hands-off dispenser”... This “beautiful thing” is following the driverless technology. Gradually mature and accelerating.

“At present, our smart cars have been able to achieve automatic driving on special sections such as low-speed traffic jams and highways. Mass production can be realized next year. After 7-10 years, it can apply to road conditions of at least 80%-90%.” Singularity Motors The founder and CEO Shen Haisong stated that the driverless driver is a technology that allows the car to have its own perception of the environment, path planning, and autonomous vehicle control. The key to this is to “teach” through artificial intelligence and deep learning. Vehicles of various types of sensors learn to drive.

From drones to network rides, from path planning to traffic management, the tremendous changes in artificial intelligence for traffic are now everywhere.

"Opening the app to call the car, the system can predict the destination based on the user's historical record, and it can pop up the recommended pick-up location." Zhang Di, chief technology officer of Didi Chuxian, took out his mobile phone and told reporters that before the passengers got on the train, Generally speaking, it is necessary to make two calls with the car driver to determine the location of the car. Artificial intelligence can record the history of car locations, track the trajectory, and determine the location of the car on the basis of experience. At present, there are 30 million such sites in more than 400 cities.

One of the technical cores of the Didi platform, Smart Dispatch, is also ushering in significant changes brought about by artificial intelligence. When passengers drove about a few years ago, the system dispatched the list mainly to consider the location, but the location did not recently mean that the pickup time was shortest. Therefore, after that, the dispatch list was added to the route planning, estimated arrival time, and the owner's service points. Multi-dimensional calculation of the matching of passengers and owners. To this day, drop orders are even smarter: considering the supply and demand of the platform and the actual situation of the road, the orders and vehicles will be blended every 2 seconds to make a global optimal match.

“Artificial intelligence has enabled path planning to change from rule-based to data-based.” Dong Zhenning, vice president and chief human resources officer of Gaode map, told reporters that the most difficult route planning is the calculation of road weights. Before 2013, they were mainly Judging from the physical conditions such as road distance, and after the introduction of artificial intelligence, it is possible to in-depth iterate through the massive data of the user's trajectory and continue to iterate, thus making the path planning more efficient and the time prediction more accurate. Using the same principle, the accuracy of forecasting of Gottel's estimated arrival time has also greatly improved in recent years, and it has reached 80% to 90%.

At the same time, companies such as Gaode and Didi are also actively cooperating with Wuhan, Jinan, and Hangzhou to optimize the traffic lights on some sections of the city. In the past, the intersections of traffic lights were generally fixed at a fixed interval. After artificial intelligence had real-time perception of traffic, the interval could be dynamically adjusted so that when the upstream convoy went downstream, the junction passed the green light. Taking Dripping as an example, as of December 10, 2017, more than 800 traffic lights have been optimized throughout the country, making the peak transit time 10%-20% lower than before.

Traveling greener, greatly improving the utilization efficiency and safety of transportation resources

“What is congestion? It is the imbalance in the utilization of road resources. It is no use to solve congestion, relying only on people’s experience, and relying on multiple roads. The best way is to use artificial intelligence to dispatch and realize people, vehicles and roads. Collaboration to improve the utilization efficiency of road resources, said Dong Zhenning.

Artificial intelligence makes travel more efficient and also makes it easier and greener.

"Good autopilot technology will reduce people's fatigue, improve the driving experience, and improve safety." Shen Hai said. Some experts said that when a person drives a car, it takes 1.2 seconds from seeing the emergency situation to step on the brakes until the brakes work. The entire reaction process of the unmanned vehicle takes only 0.1 to 0.6 seconds. The emergency brake is faster and the driver is also driving. safer.

"When the shared travel network is well-developed, it is possible to use fewer cars to meet the travel needs of more people and to greatly reduce the consumption of energy and resources." Zhang Bo said that the biggest asset of Didi is big data, artificial intelligence, Machine learning algorithms, "We want to integrate real-time traffic information and make real-time, intelligent decisions through large and complex algorithms and computing platforms."

Golf Gift

Golf Gift,Golf Gift Set,Best Golf Gifts,Cheap Golf Gifts

Shenzhen New Gaopin Sports Goods Co,Ltd , https://www.newgaopin.com