Edge Devices¶

Consystence edge devices run on Nvidia Jetson Orin modules deployed at remote substations and equipment rooms. They bridge the gap between PLCs on the plant floor and the site server on the control room network.

Each edge device handles three core responsibilities:

• PLC communication — reading and writing tag values via EtherNet/IP
• Edge ML inference — running TensorRT models for anomaly detection
• Store-and-forward — buffering data during network outages

In this section¶

• Nvidia Orin Setup — hardware specs, JetPack installation, network configuration, and .NET setup.
• PLC Communication — EtherNet/IP via libplctag, CIP tag paths, and validated PLC compatibility.
• Edge ML — TensorRT autoencoders, anomaly detection, and the fleet learning loop.
• Offline Operation — SQLite buffering, gRPC-over-HTTP/3, and air-gapped deployment.