Sensor technology for AI-supported quality monitoring in production technology

From research

Sensory tool inserts with 13 measuring points for injection molding, injection molding sample and plastic granulate.
© Fraunhofer IST
Sensory tool inserts with 13 measuring points for injection molding, injection molding sample and plastic granulate.

In order to be able to monitor production processes automatically with the aid of AI, innovative sensor systems are required that provide, with the highest possible data quality, real-time information regarding the status of the product and process. In the “AI-NET-ANIARA” project, the Fraunhofer IST therefore conducted work on the development of innovative thin-film sensors for automated production processes using the example of plastic injection molding.

Das 9. und 12. Ziel für nachhaltige Entwicklung der UN: Industrie, Innovation und Infrastruktur und verantwortungsvoller Konsum und Produktion.

The potential of thin-film sensor technology

The application of the thin-film sensors developed at the Fraunhofer IST (see figure below left) in combination with AI opens up the technological prerequisites for the implementation of autonomously controlled systems. Human operators thereby receive support in recognizing the product status within the production process and in initiating the available options for optimization and control measures as required. 

Structure of the thin-film sensor 

At the Fraunhofer IST, a multifunctional thin-film sensor system (see figure above) was deposited on steel inserts by means of physical and chemical vacuum deposition (PECVD), which can be easily integrated into the tool. The base layer consists of a thermoresistive and wear-resistant diamond-like carbon (DLC) layer. An array of 13 electrode structures made of chromium was structured on top, mapping the flow front along the component geometry. This is followed by two electrical insulation layers made of SICON®, between which the chromium-based conductor tracks were produced using photolithographic processes.

Schematic representation of the multifunctional layer system
© Fraunhofer IST
Schematic representation of the multifunctional layer system
Exemplary temperature profile over the component geometry with selected measuring points.
© Fraunhofer IST
Exemplary temperature profile over the component geometry with selected measuring points.

Outlook: Application of the sensor system in incremental manufacturing

The developed sensor system was tested on the injection-molding machine at the Institute of Machine Tools and Production Technology (IWF) at the TU Braunschweig. A temperature profile is shown exemplarily in the figure above right. As a sub-process of incremental manufacturing, digitalized and intelligent manufacturing strategies for the efficient manufacture of functionalized products in differing quantities were investigated with the aid of sensor technology. It was found that as a result of the increased data availability in combination with the application of machine learning processes, individualized (intermediate) product states can be predicted and suitable optimization strategies can be derived.

The project 

Within the framework of the EU research program AI-NET, research is being conducted into technologies that will accelerate the digital transformation in Europe. In a number of industry-led projects, the technology fields of communication networks and technologies for 5G and, prospectively, 6G, user-oriented data centers and artificial intelligence (AI) are being addressed. The German project consortium of “AI-NET-ANIARA” focused on the application fields of sensor systems and production technologies.

The “AI-NET-ANIARA” project was funded by the German Federal Ministry of Education and Research (funding number 16KIS1275) and is part of the EU research program AI-NET.

This article is part of the Annual Report 2023.

 

Annual Report 2023

We offer solutions for your applications

 

Industry solutions

Digital economy