[OSPP] Using KubeEdge to implement deep learning-based device fault detection scenarios

[Catherine-monk]

What would you like to be added/modified:
1.Implement a demo for the current scenario based on the KubeEdge framework.
2.Submit the demo application and documentation.

Why is this needed:
Edge computing, as the last mile of artificial intelligence, has extensive applications in combination with AI. In the field of industrial manufacturing, edge AI plays an increasingly important role.
Scenario: In the manufacturing process, sound detection is conducted on tools, stamping equipment, and other devices. The years of experience of skilled workers in listening to sounds is transferred to the algorithm model, allowing timely detection of equipment failures and improving the efficiency of equipment inspection and maintenance.
1.Apply the fault detection model to edge nodes using KubeEdge, ensuring that the resource consumption of the model application does not exceed 4 CPU cores and 8 GB of memory.
2.Execute the model application to collect sound signals from simulated Modbus devices. The model application should be able to identify device faults based on the sound signals, control the on/off state of simulated edge light bulbs accordingly, and record videos.
3.(Optional) Integrate terminal devices into the platform for centralized management.

refer：
https://github.com/kubeedge/examples