kaiwaehner/hivemq-mqtt-tensorflow-kafka-realtime-iot-machine-learning-training-inference
Real-time streaming ML pipeline that trains and runs TensorFlow models on IoT sensor data from tens of thousands of devices for predictive maintenance.
Velocity · 7d
+0.2
★ / day
Trend
→steady
star history
This project demonstrates machine learning at scale on IoT data using TensorFlow, Apache Kafka, and HiveMQ MQTT broker. It trains and runs inference on ML models for anomaly detection and predictive maintenance in connected car infrastructure. No external data stores are required—models are trained and served directly on the streaming pipeline processing millions of messages per second from up to 100,000 IoT devices.