Edge AI and Industrial Automation on the NVIDIA Jetson Orin Nano — a learning repository exploring real-time inference, sensor integration, and .NET-based monitoring dashboards.
Build practical, end-to-end Edge AI and Industrial Automation projects using:
- Python — inference pipelines, sensor drivers, Ollama LLM integration
- .NET / C# (Blazor) — monitoring dashboards and administrative interfaces
- TensorRT / ONNX — model optimization for on-device inference
- Docker — reproducible, containerized deployments on the Jetson
- OPC-UA + MQTT — industrial protocol integration (asyncua, aiomqtt, Mosquitto)
- PLC / Modbus — reading field device data from industrial controllers
jetson-edge-ai/
├── docs/
│ ├── adr/ # Architecture Decision Records
│ └── setup/ # Jetson and local dev environment guides
├── projects/
│ ├── visual-inspection/ # Camera-based defect detection (Python + TensorRT)
│ ├── sensor-pipeline/ # I2C / SPI / GPIO data ingestion (Python)
│ ├── anomaly-detection/ # Statistical anomaly detection on sensor streams (Python)
│ ├── llm-inference/ # On-device LLM with Ollama (Python)
│ ├── monitoring-dashboard/ # Real-time monitoring UI (.NET Blazor)
│ └── industrial-gateway/ # OPC-UA client + MQTT publisher bridge (Python)
├── shared/ # Utilities shared across 2+ projects (added when needed)
├── infrastructure/
│ ├── docker/ # Container definitions per project
│ └── scripts/ # Provisioning and deployment scripts
└── models/ # Model configs and conversion scripts (weights excluded)
| Project | Description | Stack |
|---|---|---|
| visual-inspection | Real-time defect detection via camera | Python, OpenCV, TensorRT |
| sensor-pipeline | Sensor data ingestion and streaming | Python, I2C/SPI/GPIO |
| anomaly-detection | Outlier detection on sensor time-series | Python, statistical models |
| llm-inference | Local LLM chat and RAG experiments | Python, Ollama |
| monitoring-dashboard | Live metrics and device management UI | .NET 10, Blazor |
| industrial-gateway | OPC-UA client + MQTT publisher bridge | Python, asyncua, aiomqtt |
See docs/setup/jetson-setup.md for Jetson Orin Nano provisioning and docs/setup/development.md for local development environment setup.
- Device: NVIDIA Jetson Orin Nano
- JetPack SDK: 6.x (L4T r36)
- Documentation: Jetson Orin Nano Developer Guide