The edge cloud cooperation platform can provide a wide range of application scenarios by integrating a variety of training inference algorithms suitable for the edge. Also provides cloud collaborative library code, compatible with mainstream frameworks to simplify the development process. In addition, the platform optimizes communication and storage for edge cloud collaboration, and improves the efficiency of training reasoning. What are the specific implementation methods of edge cloud collaborative platform in industrial applications?

End-to-end field iot ubiquity awareness construction. Through submarine optical cables, the optical fiber loop of offshore oil field groups is built, and 19 types of data such as personnel status, equipment status, platform and surrounding environment, and production process are collected and aggregated. The use of WIFI6, 5G and other communication technologies and industrial network equipment to achieve wireless network coverage in the production site, network access in key areas, and open up the iot nerve endings that sense the access of equipment.

Edge-side edge cloud construction. Build edge micro-clouds on offshore platforms and deploy edge applications containerized. Through edge technology to achieve on-site data collection and storage, local computing, reduce the pressure on network transmission and cloud resource consumption, to meet the application requirements of real-time business.

Training and development of cloud models. The central cloud conducts model training based on Keras. After the training is completed, the algorithm is deployed to the edge node to predict new images in real time, store and upload the prediction results. Through continuous model training and result feedback, iterative updates are made to form a systematic closed-loop and deliverable application model under cloud edge collaboration.

Data acquisition and preprocessing. Through the collection of picture data, the data is further encapsulated and the data transmission speed is improved. To realize data synchronization or reception, through the establishment of data channels in the edge nodes and central cloud nodes, real-time transmission of all kinds of data, reception of received data, and storage and processing according to demand, further providing data foundation for applications.

Model training and optimization, through the traditional industrial technology principles, industry knowledge and experience and production attacks, the implementation of deep learning and machine learning algorithms to achieve artificial intelligence applications in industrial scenarios.

Business orchestration coordination, edge nodes support modular, microservice application instances, the cloud provides the ability to achieve business orchestration according to customer needs.

By integrating IT and OT systems, IT can collect edge device data, connect to enterprise ERP, PLM and other service systems, and deliver control commands and configure parameters to achieve integration and application development between IT and OT systems.

In the intelligent edge platform, visual management of independently distributed individual servers and service differentiation are used in the system, providing automated operation and maintenance capabilities such as mass, automatic troubleshooting, upgrade, and service operation, reducing the difficulty of operation and maintenance for non-professional IT staff, and realizing automated and streamlined operation and maintenance.

Support cloud network integration application, direct to the industrial site cloud network integration edge intelligent scene solution, providing end-to-end turnkey services.

The above is the edge cloud collaborative platform can achieve data processing and analysis in industrial applications, improve the efficiency and quality of industrial production processes, and promote the digital transformation and intelligent upgrading of industrial enterprises.