View-driven Real-time Sensor Data Processing and Servitization System

In the era of the Internet of Things, the amount of data generated by industrial sensors is large and dynamic, and its processing requirements are also complex and variable, requiring developers to customize the algorithm, the process is cumbersome and the development cycle is long. In this paper, a real-time processing system of sensor streaming data based on data flow view is designed. The data stream processing process is abstractly encapsulated and customized assembly based on the minimum processing unit to form a complete processing logic link. The view logic is mapped to the processing function at the bottom of the distributed system through the process parsing engine to form a task list, which drives the stream data processing process based on the Kafka message queue and the Flink parallel computing framework. The processing result is pushed to the application system in real time via the service route to meet different business requirements. The case study shows that the system has flexible and configurable basic functions of stream processing and has good horizontal scalability.