China's application of transportation big models to improve transportation efficiency and safety
2023-09-28
"Traffic is a complex system covering perception, cognition, decision-making and control, and the key to breaking the problem is to apply the big model to this complex process." At the Global Sustainable Transport Summit Forum held in Beijing on the 25th, Shen Shuan, executive vice president of Baidu AI Cloud Group and chief executive of Baidu Intelligent Cloud Business Group, said. On September 21st, in the indoor exhibition area of the 2023 World Intelligent Connected Vehicle Conference, two visitors were visiting Xiaopeng Flying Car. Ju Huanzong, a journalist from Xinhua News Agency, pointed out that traditional artificial intelligence can only achieve fragmented point like applications in the transportation field, such as identifying a traffic light or a certain violation behavior. Shen Dou said that in order to truly solve the complexity and coupling problems of the transportation industry, it is necessary to leverage the value of large models by dynamically integrating data information from vehicles, roads, clouds, maps, and other aspects to improve transportation efficiency and safety. The transportation big model is an industry big model developed based on a universal big model. For the large model of the transportation industry, first of all, it is necessary to provide professional knowledge in the transportation field to the large model. Secondly, the case studies, practices, and expert experience in the transportation field should be compiled into training materials to enable the large model to continuously learn. At the same time, the model should be brought into industry practice to continuously verify and provide guidance and optimization for problems. "Sun Yafu, Vice President of Qianfang Technology, introduced. The impact of digital technology on the transportation industry is extensive and profound. In the process of China's transformation from a transportation power to a transportation power, Baidu AI Cloud, Qianfang Technology and other domestic cutting-edge technology enterprises have promoted the coordinated development of vehicles and roads in multiple pilot cities and the construction of urban intelligent network connectivity with the mode of "unveiling and leading". Currently, China is promoting the digital transformation of the transportation industry through technologies such as cloud, graphics, big data, and digital twins. Among them, the transportation big model has been applied in multiple scenarios such as China's intelligent network, intelligent highways, and urban transportation. On September 21st, in the outdoor exhibition area of the 2023 World Intelligent Connected Vehicle Conference, visitors visited the display of autonomous small buses. Xinhua News Agency reporter Ju Huanzong took a photo in Jiading, Shanghai. More than 800 autonomous vehicles have been launched with over 30 intelligent applications, providing intelligent services such as traffic light countdown, dangerous intersection prompts, and path planning for car owners, with an accuracy rate of over 99%. In terms of urban transportation, the transportation big model can use global perception data to analyze and judge the possibility of congestion occurrence, and overlay simulation deduction to generate a global signal control plan. For example, a traffic model can predict where, why, and how to deal with traffic congestion at 11 o'clock at 9 o'clock, and automatically optimize the signal timing plan. Baidu researchers told reporters that after laying out a comprehensive signal control plan for a 60 square kilometer area in Yizhuang, Beijing, the average passing time at intersections has been reduced by 28%. Recently, the highway AI digital person "Jian Lulu" appeared on the Hebei Expressway, providing a new possibility for the digital development of highways. Digital humans can not only detect road accidents such as rear end collisions, but also automatically search for nearby rescue resources and generate accident analysis reports. After application, the efficiency of on-site event handling will double and labor costs will be significantly saved, "said Baidu's research and development personnel. With the increase in training volume, "Jian Lulu" has now served in road network monitoring, emergency command, maintenance management, and public transportation
Edit：Hu Sen Ming Responsible editor：Li Xi
Source：XinhuaNet
Special statement: if the pictures and texts reproduced or quoted on this site infringe your legitimate rights and interests, please contact this site, and this site will correct and delete them in time. For copyright issues and website cooperation, please contact through outlook new era email：lwxsd@liaowanghn.com
Sci-Tech
Recommended Reading Change it 
Home
