In April 2016, in a speech on the 5G Internet of Vehicles Demonstration Project in Yunqi Town, Hangzhou, Wang Jian, the "Honorary Mayor" of Yunqi Town, founder of Alibaba Cloud, and Chairman of the Alibaba Group Technical Committee, proposed the "Urban Data Brain". The concept was pioneered, and it was immediately resonated and highly recognized by city managers and relevant government departments. Two months later, the Hangzhou City Data Brain Project was launched as an exploratory project to improve urban traffic congestion.
At the Yunqi Conference in October of that year, the Hangzhou Municipal Government, Alibaba Group, Dahua Technology Co., Ltd., China Control Information and other 13 units jointly officially released this huge project. Wang Jian's proposal at the Yunqi Conference that "the farthest distance in the world is from the camera to the signal light" has been highly hyped by the technology media and more "denounced" by the intelligent transportation industry.
In March 2017, the concept of "urban data brain" came out of Hangzhou, and then Quzhou, Wuzhen, Suzhou, Tongxiang and other cities began to build "urban brains" led by Alibaba Cloud.
At the Yunqi Conference a year later, we heard the latest progress in the construction of Hangzhou's "Urban Data Brain" project. The work of "Internet + Signal" has been tracked, and we will look back to see how much work has been done on the "farthest distance", which has caused great controversy, and how the technology of the intelligent transportation industry has developed and progressed.
1. What is the city brain
The urban brain has a lot of connotations, but the current focus is on the depth | "Urban traffic brain" makes more technological progress in intelligent transportation, and the "traffic brain" in its branch transportation field is the one with the most demonstration construction and technological exploration.
Alibaba Cloud defines the "city brain": the city brain is a brand-new infrastructure that supports the sustainable development of future cities. Its core is to use real-time full city data resources to globally optimize city public resources, instantly correct city operation defects, and achieve the following three: a breakthrough:
Breakthrough in urban governance mode: Improve government management capabilities, solve outstanding problems in urban governance, and realize intelligent, intensive and humanized urban governance;
Breakthrough in the urban service model: more accurately serve enterprises and individuals anytime, anywhere, the city's public services are more efficient, and public resources are more economical;
Breakthroughs in the development of urban industries: Open urban data resources are important basic resources, which play a leading role in industrial development and promote the transformation and upgrading of traditional industries.
Alibaba Cloud City Brain Overall Architecture
From Alibaba Cloud's interpretation of the overall structure of the city brain, traffic, traffic police, and public transportation are the main data sources and application scenarios.
In fact, the concept of "brain" first appeared in the intelligent transportation industry around 2006. At that time, the meaning of "traffic brain" was not the same as what we mean by "traffic brain" now.
At present, not only Ali is mentioning the term "transportation brain", Baidu and Didi Internet companies are also mentioning it, traditional intelligent transportation companies and industry users are mentioning it. "Transportation brain" is not a professional term, it has become a Marketing vocabulary.
II. Urban Traffic Brain Technology Route
As Internet companies gradually develop their business in the field of intelligent transportation, whether it is Internet + signal or Internet + traffic management, the entire industry is gradually showing a relatively consistent technology development route based on traditional intelligent transportation technology and the combination of Internet technology. .
Fusion of Internet Data and Fixed Point Data
How individual Internet data "runs rampant" in the field of intelligent transportation has become a false proposition, and the fixed-point traffic collection data that has been invested heavily in the project market for many years has been of limited significance for many years. However, the fusion of Internet data and fixed-point data of public security traffic police, on the one hand, uses Internet data to calibrate traditional detection data to obtain macro-continuous data; The spatial granularity and confidence of the data.
Data fusion gives full play to the macro, comprehensive, coherent, and continuous characteristics of Internet data, and then uses fixed-point data to supplement microscopically. With big data, good things will come.
We have seen whether it is the latest intelligent transportation platform products released by Yihualu, Yinjiang and Hisense this year, or the just-released Guangzhou Internet + Traffic Light Research Project, Zhongshan Internet + Smart Transportation Project key research and construction content, or The basic data of Hangzhou Urban Traffic Brain is the fusion of Internet + fixed point data. This is one of the important foundations for future traffic management, and it is also one of the entry points for the cooperation between Internet companies and traditional intelligent transportation companies.
Transportation data center establishment
I have always believed that the so-called structuring or "whitewashing" of traffic data is the most critical link and technology in the application of all big data technologies, followed by various core algorithms.
One of the construction contents of the Zhongshan Internet + Smart Transportation Project is the construction of a traffic data computing center: access to Internet traffic data and traffic management-specific data to form a standardized and spatialized resource catalog.
In the Xiaoshan District Demonstration Project of Hangzhou Urban Traffic Brain, two of the three main contents involved data application. One is the construction of the Public Security Cloud Eye, which conducts structured analysis and processing of more than 800 videos at intersections and roads; the other is the construction of a data resource service platform, which takes data as the core to realize the construction of a resource platform with the main goals of data processing, development and service. , including a resource platform for data collection and precipitation, data governance, data catalog, data engineering, data management and control, and data services.
We often hear that urban traffic surveys, a large amount of traffic detection data are sleeping in the computers of relevant departments, and the utilization rate is low. We also heard that an Internet company cooperated with a city in an Internet + signal optimization project. When integrating with fixed-point data, the structuring of fixed-point data is assisted by the Internet company, and users do not know how to deal with the massive data.
Now, the construction of traffic data centers, traffic data platforms, and the emphasis on technical aspects make it possible for data resources to become more valuable.
Internet + Signal is one of the core applications
As the core of traffic management, scheduling and service, traffic signal control has become the core component of "traffic brain", and Internet + signal has become one of the core applications of "traffic brain".
Hangzhou City Brain currently conducts relevant research and practice work around traffic signal control.
The "Internet + Signal Light Control Optimization Experimental Research Project" in Guangzhou is actually researching the "switching signal light control optimization method". The control parameters of intersection signal lights are relatively complex, while the control parameters of tidal lane lights, ramp control lights, and sub-lane lights are simple, and only involve on and off time and duration. The Guangzhou project focuses on the research on the control method of switch-type signal lights. By demarcating the area scope of the control impact, correlating the front and rear control sections, and conducting real-time traffic estimation and impact assessment, the on-off and switching models of the signal lights are determined.
The fifth construction content of Zhongshan's "Internet + Intelligent Transportation Project" is "Internet + Signal Control Evaluation and Optimization Platform". Combining the Internet traffic data and its feedback information with the signal control system, establish a continuous signal control optimization strategy, and form an optimization plan for the control of unreasonable single-point intersections with unreasonable signal timing. The average road speed provides a more accurate phase difference for the coordinated control of the trunk line, and realizes the "variable green wave band" of vehicle traffic; the road network topology relationship provided by the traffic geographic information service system and the vehicle trajectory data of the traffic data calculation center are combined with the regional Constitute the intersection congestion detection results, and form regional traffic flow induction and signal optimization plans.
Data visualization, data-driven traffic management
After the processing of big data, the monitoring, judgment and display of traffic conditions have become an indispensable part of business applications.
The sixth construction content of the "Internet + Intelligent Transportation Project" in Zhongshan City, the construction of the traffic management visual command platform requires the realization of traffic operation overview monitoring, Zhongshan City traffic road network road condition monitoring, regular congestion/abnormal congestion monitoring, hot spot area operation status monitoring, Traffic incident monitoring and release, police resource monitoring and display, accident and illegal monitoring, spatial display monitoring of traffic facilities and equipment, video inspection and other functions. Realize the research and judgment of traffic management big data. Through the aggregation and integration of Internet traffic travel data and data of various business subsystems of traffic management, it can realize quantitative research and judgment analysis of urban road network operation index, incident information judgment and alarm, visual display of situation evolution, regular congestion and abnormal congestion. Analysis, event impact analysis, command index research and judgment, traffic OD and vehicle big data migration map, etc.
In the urban brain demonstration construction project in Xiaoshan District, Hangzhou, the construction of the brain operation command platform includes the construction of a traffic management visualization application large screen, which displays the overall traffic situation in real time, real-time signal timing, real-time traffic prediction, traffic anomaly feedback and other data application results. .
System performance evaluation method
In the application of Internet data in intelligent traffic management, it seems that the biggest use at present is to monitor, evaluate and predict the operation of urban road networks and road sections. More and more people agree with this point of view, although Internet companies think they More can be done.
In the huge controversy, the evaluation indicators of the Internet + signal optimization effect are often questioned, and most of these indicators are evaluated and released by Internet companies and are not transparent.
In the "Internet + Intelligent Transportation Project" of Zhongshan City, it is required to build an "Internet + Signal Control Evaluation and Optimization Platform", based on the Internet data, the real-time monitoring of the entrance lane speed, travel time, and queue length data at each signal control intersection in Zhongshan City, and Can be displayed visually.
Guangzhou's "Internet + Signal Light Control Optimization Experiment Research Project" requires statistical mining of traffic flow characteristics based on the fusion and processing of fixed point data and Internet data, and estimation of traffic parameters, including queue length, parking times, Travel times, entry and exit speeds, flow estimates, etc., are parameters that apply an evaluation model for the operation of the intersection road segment.
The urban brain evaluation system in Xiaoshan District, Hangzhou is more complex. For the first time, six mutually corroborating models including questionnaire survey, telephone survey, vehicle commuting time calculation, road operation data evaluation, traffic change comparison at frequent congestion points, and actual road driving experience are introduced. Signal control evaluation.
The evaluation method of transportation system operation is the key to evaluating the effectiveness of Internet + transportation-related technology application work, and the research and rigor of its evaluation method are particularly important.
Source: Saiwen Traffic Network
