In May 2018, the executive meeting of the State Council decided to cancel the provincial toll station of expressway. One year later, in May 2019, the State Council issued the Implementation Plan for Deepening the Toll Road System Reform and Canceling the Provincial Toll Stations of Freeways, officially pushing the toll road system reform to the fast lane. The Scheme requires that the ETC penetration rate will reach more than 90% by the end of 2019, and mobile payment will be fully covered in the manual toll lane. The digital transformation of expressway has begun.

While the public has experienced cost saving and convenient passage, the management and operation units of expressways are facing difficulties under new circumstances: some undesirable elements use the "loopholes" of the system to evade fees, and the ways of fee evasion are diversified, including card changing and fee evasion, front trailer separation and fee evasion, switching electronic tags, ETC lane following and fee evasion, etc, At the same time, the behavior of evading fees has evolved into specialization and gangs, which has brought a lot of economic losses and serious challenges to high-speed operation units.
At present, the toll gate still mainly uses the traditional audit method. In traditional ways, preliminary abnormal vehicle driving information (such as frequent entry and exit of the station) is usually taken as the basis for screening, and a large number of evasion behaviors are difficult to be found in time.
The main problems of traditional methods are:
There are few big data analysis applications and lack of in-depth data analysis;
Mainly based on charging data, with insufficient supporting evidence such as video/pictures;
It is inefficient to rely on a large number of manual audit.
In view of the multiple pain points of traditional audit methods, Huawei Cloud officially launched the highway big data audit solution.
This scheme is based on Huawei's leading cloud digital platform in the cloud industry, combined with Huawei's deep accumulation in the highway industry, and using advanced technologies such as big data, artificial intelligence, cloud side collaboration, etc., to realize the automatic analysis of massive traffic record data, and combined with the image records captured by the gantry camera, to achieve a precise attack on vehicles that evade fees and ensure the benefits of high-speed owners.

Huawei Cloud Big Data Audit Solution includes three platforms:
AI Edge Audit Platform Based on Huawei's self-developed Kunpeng 920 and Shengteng 310 chip+intelligent edge platform IEF Edge architecture, real-time processing of 30+vehicle characteristics and identification and audit scenarios of tens of thousands of vehicle models, as well as the storage of vehicle passing photos are realized;
Big data audit platform , mainly including infrastructure layer, platform layer, enabling layer and application layer:
Infrastructure layer: provide computing, storage, network and other basic resources
Platform layer: provide general platforms and components such as intelligent data lake platform DAYU, AI development platform ModelArts, edge management platform IEF, database and middleware
Enabling layer: provide basic capabilities such as vehicle identification algorithm, image search, path restoration, etc
Application layer: including audit system, customer service system and other public service systems. The audit system mainly implements audit related functions such as evasion fee model, access record analysis, evidence chain management, credit management, black and white list management, etc
Vehicle feature training and development platform To realize the continuous improvement of new vehicle type recognition and vehicle feature recognition capability, and the permanent storage of abnormal vehicle traffic photos; Among them, ModelArts is a one-stop AI training and development platform, providing massive data pre-processing and semi-automatic annotation of vehicle models and vehicle characteristics, large-scale distributed training, automatic generation of vehicle recognition models, cloud side on-demand deployment models and other capabilities.
The main advantages of the scheme are:
Unified management of edge equipment and business, Unified batch distribution of algorithm models , greatly simplifying system operation and maintenance.
In the project of canceling the provincial toll station, the relevant national departments put forward clear requirements for network security: fully implement the third level requirements of hierarchical protection.
Huawei Cloud is committed to providing customers with a highly reliable business operation environment, easily accessible, on-demand, and elastically expanded cloud security services to help customers protect applications and important data on the cloud. Huawei Cloud has obtained more than 20 security compliance certifications such as CSA STAR and ISO security systems, and passed the four level evaluation of equal protection with high scores in 2018. Huawei Cloud should not only be used well by customers, but also be used with confidence.

Each highway connects the whole country and all corners of the world, like a strong artery running across the skin of China, providing endless power for the operation of society. The cancellation of provincial toll stations and the popularization of ETC mark the deepening of intelligent transportation reform, making the connection more close and smooth. Huawei Cloud big data audit plan hopes to help customers to promote the refinement of highway management and provide more convenience for public travel.