README.md

IoT Predictive Maintenance using ML

In this scenario, we illustrate the example of using machine learning and IoT data pipeline to support predictive maintenance.

ML and other services/units

We use ML Units for BTS Prediction for:

• create ML services by deploying ML models into a service
  • the service accepts requests from messaging systems (currently using AMQP)
• using other units/services for emulating sensors
• ML clients can take data from sensors, preparing data for ML requests and sending the requests to the ML service

  • Currently a ML client just reads pre-processing data from files. A developer can change the code to read data from MQTT (e.g., using our IoTCloudUnits) and to perform the data preprocessing