As part of beyerdynamic’s Industry 4.0 mission to continuously optimise production, the company has introduced state-of-the-art production lines. In addition to using various high-tech robot arms to carry out production steps, detailed quality checks are performed on the workpieces.
As the results depend on a number of different machine control parameters and the production lines, in turn, produce a number of different systems, determining the optimum configuration of the input parameters in each case is incredibly complicated.
In this, and in other complex tasks, therefore, beyerdyamic is supported by highly efficient, real-time Data Science algorithms.
Demand-based computing power from the Microsoft Cloud
Because real-time processing requires an unpredictable amount of computing capacity, beyerdynamic decided to outsource large portions of their data processing and storage needs to Microsoft’s public cloud. inovex played a leading role in supporting the company through the design and implementation of the solution.
The infographic shows the essential components of the solution architecture and clarifies their interactions. The grey area on the left side shows the new production line (including the quality control station) and the field gateway which serves as the central collection and transmitting area. The rest of the infographic shows the services running in the Microsoft Azure cloud.
Azure Stream Analytics, Machine Learning & Data Lake
While the workpieces run through the steps in the production line, the input parameters used are transferred to the field gateway. This data and the inspection reports are sent to the Azure cloud using an encrypted connection and specialised IoT protocols.
The various data streams are consolidated in Azure using the Azure Stream Analytics component, which enables real-time analysis using user-friendly SQL syntax. From here, the data is transferred directly to various services, including Azure Machine Learning (where data science models calculate optimisations), Power BI, which monitors current production, and the Azure Data Lake for comparison with historical data.
The solution is enhanced by functions which write specific results back to beyerdynamic’s on-premise devices by means of Azure Functions and the IoT Hub.
Platform as a Service – ideal for SMEs
All the components for the Microsoft Azure scenario fit with the Platform-as-a-Service concept (PaaS). This means that the cloud provider is responsible for the operation and availability of the infrastructure and guarantees certain service levels (SLAs).
It also eliminates the need for major investments in proprietary data centres, and keeps the operating costs calculable and manageable. In addition, there is no need for beyerdynamic to have an encyclopaedic knowledge of the new technologies. The cloud solution also provides almost endless resources which ensure long-term demand-based scalability, even with dramatic increases in data volume.
Related projects use the Cognitive Services for speech recognition, additional Azure data services to synchronise data across locations, and Power BI components for dashboards. Microsoft Power Apps are also used intensively in quality control.
- Azure IoT SdK
- Azure IoT Hub
- Azure Stream Analytics
- Azure Data Lake Store
- Azure Data Lake Analytics
- Azure Machine Learning
- Azure Functions
- Visual Studio
- Power BI
The inovex team and the Microsoft Cloud allow us, as an SME, to use highly innovative, leading-edge technology to optimise our business processes. The inovex team aren’t just experts, they’re also thoroughly nice people.