IoT Data Platform#

Manufacturing processes generate a wealth of data. As more and more processes are automated, the equipment becomes more sophisticated, and sensors become widely used, the amount of data that will be generated in the shop floor will grow significantly. Also, even though there is data that is valuable today, there is data for which there is no known usage currently, but that may be very useful in the future. A key requirement in any manufacturing environment is the capability to store massive amounts of data and to store it for a long period of time. A second key requirement is to be able to process and analyze the data, either online (stream processing) or offline (batch processing). These types of analysis are useful to get actionable results quickly. These actionable results can be carried out automatically by the system, for example, by triggering a maintenance request or manually if the system is used only for decision support. The IoT Data Platform application is a highly scalable platform based on Apache Kafka™ and Apache Spark™ to ingest, store and process events. The system is based on a meta data registry, that specifies the different system events, as well as their routes to different event consumers. The system provides out of the box a consumer to write the event data into CSV/JSON and SQL Server, but given its flexible and extensible architecture, new consumers can be added seamlessly to the platform.

Data for the IoT Data Platform can be provided by Connect IoT or from any other application using the ingestion layer APIs. Given the wealth of data that can be ingested, processed and stored in the IoT Data Platform, in combination with the data available from the MES, it’s possible to create all sorts of models, analysis and algorithms that can be used for real-time process control as well as for yield improvement and root-cause analysis.

Available guides#

Table: IoT Data Platform guides