The advantages of Sensor-Intelligence Devices (SID)
Real-time, energy-optimized sensor intelligence devices (SID), such as those developed by the Fraunhofer Center for Sensor Intelligence ZSI, act as an intermediate or meta-level between operational technology (OT) at the machine level and IT systems at the business level: for example, a manufacturing execution system (MES), an IoT platform, or an ERP.
To establish interoperability, i.e. linking, communication and collaboration between SID and IT systems (MES, IoT platform, ERP), and to integrate process-relevant data into business processes, an IT solution such as ORBIS Distributed Shopfloor Processing (ORBIS DSP) is needed.
Condensing sensor data into process information
From a technical point of view, SID are electronic components that combine raw data or processed data from the sensors on a machine, combine and condense it into process-relevant information and thus allow statements to be made about the manufacturing and processing quality of a product. SIDs are also used to monitor and adjust production processes in real time and to optimize them in terms of product quality, safety, infrastructure or resource utilization.
Reduction of sensor data saves costs
Usually, the condition data recorded by sensors, such as temperature, pressure or vibration, but also quality characteristics such as pores or cracks in a casting or the press hardness of a material, flow into an MES or an IoT solution, where they are correlated into business-relevant information. The result is a permanent stream of data that is energy-intensive and therefore costly, and at the same time a huge and constantly growing volume of data - big data, in other words - that requires enormously large and expensive storage capacities.
Real-time, energy-optimized sensor intelligence devices (SID) close the gap between operational technology (OT) at machine level and IT systems at business level.
Prof. Dr. Hans-Georg Herrmann,
Fraunhofer Institute for Non-Destructive Testing IZFP
The SIDs, on the other hand, condense the flood of sensor data and thus the actual values into process-relevant information, for example about the hardness of a metal, and transmit this condensed data to the business systems, which significantly reduces the data flow. This not only saves power, but also expensive storage space - a clear added value.
The potential of AI-supported sensor systems
However, the potential of SIDs is not yet exhausted: They can be interconnected to form an intelligent sensor network in which they interact in real time, and can be enhanced by artificial intelligence (AI). With the help of AI-supported sensor systems, statements about the quality of a component can be made directly in the manufacturing process, because anomalies and defects such as pores in a casting or cracks in a screw are identified immediately.
This enables adaptive process control and optimization during ongoing production, as plant or product parameters can be dynamically adjusted, but also predictive maintenance.
AI-supported sensor systems from Fraunhofer ZSI are already proving themselves in practice in various application areas. In the automotive industry, for example, they are used to inspect press-hardened car body parts and to check valve springs, in friction stir welding for real-time monitoring of process stability and joint quality, or for final assembly inspection of cutting tools for combine harvesters.
For more information on Sensor Intelligence Devices (SID), see:
