🌕 Moonshot IoT Stack

Moonshot IoT Stack is an end-to-end industrial IoT data pipeline and edge inference platform. It integrates MQTT ingestion, time-series storage, real-time dashboards, and ONNX-based ML inference—all containerized and ready to deploy.

🚀 Features

• Node-RED for visual data flow orchestration
• InfluxDB for time-series storage
• Grafana for live dashboards
• ONNX Model API for local machine learning inference
• Mosquitto MQTT broker for sensor ingestion
• Docker Compose-based for easy deployment

🧱 Stack Architecture

[Sensor] --> [Mosquitto MQTT] --> [Node-RED] --> [InfluxDB] --> [Grafana]
                                     |
                                  [ONNX Model API (Flask)]

📦 Requirements

• Docker
• Docker Compose
• Git
• Python 3.8+ (for model development only)

🛠️ Setup

1. Clone the repo

   git clone https://github.com/iuriarte/moonshot-iot-stack.git
   cd moonshot-iot-stack

2. Start the stack

   docker compose up --build

3. Access Services

   • Node-RED: http://localhost:1880
   • Grafana: http://localhost:3000 (admin / admin)
   • InfluxDB: http://localhost:8086
   • ONNX Model API: http://localhost:5001

🧪 Inference API Usage

Make a POST request to /infer with input data:

curl -X POST http://localhost:5001/infer \
  -H "Content-Type: application/json" \
  -d '{"inputs": [[0.0, 1.0, 2.0, 3.0, 4.0]]}'

Other endpoints:

• /healthcheck – check if the model loaded
• /readycheck – confirms API is alive

🧠 Model Format

Place your ONNX model in:

model/model.onnx

Update environment variable MODEL_PATH if using a different path.

🧪 Development Tips

To test just the model API:

cd model
docker build -t moonshot-modelapi .
docker run -p 5001:5000 moonshot-modelapi

Or pull from Docker Hub:

docker run -p 5000:5000 iuriarte/moonshot-iot-stack:modelapi

📁 Folder Structure

moonshot-iot-stack/
├── model/
│   ├── model.onnx
│   ├── model_server.py
│   └── Dockerfile
├── docker-compose.yml
├── config/
├── data/
└── README.md

🤝 Contributing

Pull requests are welcome. Please open issues first to discuss changes.

---

📄 License

MIT License – see LICENSE file for details.