Trend: New ICT Technologies Accelerate Intelligent Transformation

Computing Power: The rapid development of chip technology provides super - strong computing power support for AI. At the front - end, edge, and center, AI chips offer computing power support for the in - depth analysis of massive videos, pictures, and data, as well as for big - data collisions and retrievals.

Algorithms: Supported by computing power, general - purpose AI algorithms are becoming more and more mature. There is a need for long - tail small algorithms that can adapt to industry processes and have industry attributes.

Big - data Technology: It provides a high - performance processing platform for massive data and accumulates effective data assets. Through extensive aggregation and efficient governance of massive data, multi - dimensional data fusion is achieved.

Front - end: It is not just for simple data collection. Instead, it has front - end analysis capabilities according to different application scenarios.

Back - end: The system is "cloud - enabled", diluting the concept of integration and compressing intermediate links. The non - discriminatory service capabilities of the cloud make it possible for the industry to open up capabilities, interconnect services, and carry out collaborative innovation.

Platform: It strings together the "pearls" in various fields such as computing, storage, connection, cloud - enabling, intelligence, and security. 

Integration of Internet of Things, edge computing, cloud platform and AI technology to build an AIoT platform

Challenges in Application Intelligence

Low Degree of Intelligence with Extensive Manual and Off - line Processing 

With the high - definition of surveillance images, videos record a wealth of valuable information, such as object features, vehicle information, facial features, and behavior information. They play an increasingly important role in all aspects of security and evidence - collection. Moreover, video applications are gradually evolving from simple visualization to comprehensive digital and intelligent applications. To match the fragmented industry scenarios and diverse environments, a large number of long - tail small algorithms with industry attributes are needed. These algorithms are rooted in industry production processes and need to have the capabilities of rapid launch, continuous evolution, and on - demand use to solve customers' actual problems.

Challenges of Safe Operations and Environmental Protection Requirements

Urgent need for timely identification and rapid disposal through digitization

The law enforcement of safety supervision and environmental protection has been strengthened, and the cost of mandatory renovation is high: the special rectification of gas stations by local safety supervision bureaus has become normalized, and many gas stations with safety/quality hazards have been shut down, with unprecedented law enforcement efforts. At the same time, the environmental requirements for gas stations have been increasing year by year, such as the implementation of the State Council's "Ten Measures for Water" in 2015, which prompted local governments to put forward mandatory requirements for the renovation of double-layer tanks at gas stations.