Self-Service Predictive Maintenance for Leading German Car Manufacturer


Achievement

  • Technical implementation 1 year
  • Running in 3 production plants
  • Savings 4 million Euro annually

Summary

Self-service based production predictive maintenance solution was developed to enhance rapid machine learning prototyping for production engineers while keeping lower barrier from machine learning knowledge. It has support for integration with public and in-house cloud and possesses the capability of ingesting and processing high volumes of high-velocity unstructured data. The solution is currently used in three production plants of the leading car manufacturer.


Technology Used

  • Azure IoT-Hub
  • OPC UA
  • Azure Stream Analytics and Event Hubs
  • Azure Functions and Spark Databricks
  • Azure Cosmosdb and Microsoft SQL Server
  • Azure Data Lake and Blob Storage
  • Tensorflow/ Scikit Learn
  • Microsoft PowerBI
  • R, Python
  • PHP, Bootstrap