With the proposal of the Internet of Things (IoT) concept, the IoT device has gradually become the center of data production, and its production data is growing faster and faster. Furthermore, for its geographical dispersion and the increasingly high requirements for response time and security, the existing cloud computing platform has become more unsuitable for the actual scenario, and the stage center of data computing is gradually moving from cloud computing to edge computing.

The following takes the IoT scenario and AI scenario as examples to share two typical solutions.

Scenario 1 Edge IoT data analysis

In the face of a complex data collection environment, diverse data communication protocols, massive raw data, and different data flow requirements, Baidu IntelliEdge can easily establish the edge computing applications integrating with the data collection, protocol resolution, data analysis, and data forwarding through the combination of functional modules, to meet the general requirements of most IoT scenarios in industrial production and urban monitoring.

Scenario 2 Edge AI machine vision

Regardless of face recognition, industrial quality inspection, urban management, or public safety, the machine vision model that completes training in the cloud has been applied to all walks of life. The BIE cloud management suite is easy to deploy such visual models to the edge node. On the one hand, it can provide a fast local recognition response through the edge AI inference. On the other hand, it can reduce the transmission bandwidth cost of videos/images through the edge AI inference.