University of Bremen: AI is revolutionizing the condition monitoring of machines!

University of Bremen: AI is revolutionizing the condition monitoring of machines!

In the “MasterKI” research project, the University of Bremen has committed itself to developing an intelligent condition monitoring system for mobile work machines. Partners in this innovative project are ANEDO GmbH, SEGNO Industrie Automation GmbH and the ITEM research institute at the University of Bremen. The goal is to create a modular edge solution that monitors machine health in real time using an AI-powered cloud platform. This solution is particularly important for mobile machines such as harvesters and straddle carriers, which are exposed to high loads during their operations. Loud uni-bremen.de Current monitoring methods are often costly and provide limited data. The “MasterKI” approach is intended to enable flexible and scalable condition monitoring through the use of edge computing and artificial intelligence.

Professor Karl-Ludwig Krieger highlights the challenges associated with developing a robust and adaptable system. A central component of the project is a cloud-based platform responsible for signal pre-processing, condition monitoring and data transformation. The plan is to close the gap between test bench data and data from real-world applications. Julia Scholtyssek, a project participant, explains that transfer models and machine learning help reduce the dependence on extensive field measurement data. The entire research approach is supported by the Federal Ministry for Economic Affairs and Climate Protection (BMWK).

Focus on goal setting and innovation

The integrated edge measurement system and the cloud-based analysis platform are intended to enable reliable monitoring of the drive units of mobile machines. Matthias Terhaag, project manager at ANEDO, emphasizes the importance of identifying potential damage early in order to prevent costly failures. Data security also plays a central role, especially since mobile machines are often used in safety-critical areas. The solutions under development use modern encryption technologies to prevent unauthorized access. A user-friendly app is being developed to optimize the control and monitoring of the systems. Vasco de Freitas from SEGNO emphasizes that the solution not only increases machine availability, but also significantly reduces operating costs.

In addition to the work of the University of Bremen, Fraunhofer IPMS has created a demonstrator based on the results of the iCampus project ForTune. This demonstrator combines sensors, data acquisition and AI-supported data evaluation to guarantee precise condition monitoring and predictive maintenance of machines. Dr. Marcel Jongmanns, who leads the project, emphasizes that the integration of AI into the sensors makes it possible to detect damage early and optimize maintenance intervals. This technology will also be presented at the upcoming SENSOR+TEST trade fair from June 11th to 13th, 2024 in Nuremberg. A miniaturized conveyor belt is presented in a show case that illustrates the possibilities of industrial plant monitoring.

Technological advances and their application

Multimodal sensors that record accelerations, rotation rates, magnetic fields as well as acoustic and ultrasonic signals form the basis for the system. The system solution also enables the integration of in-house sensors with an edge computing unit based on the RISCV architecture. This allows complex AI operations and real-time analysis directly at the point of use. Real-time calibrations increase the accuracy of the models and adapt the system to new environmental conditions. The continuous development of the technology is also supported by existing partnerships, for example with Vetter Kleinförderträger GmbH, which illustrates the interest of the industry.

Given the extensive data processing that is necessary for precise damage prediction, the importance of data understanding and processing is highlighted. This is a key area that requires a solid data strategy in addition to pure technology development. External support is often required to adequately assess opportunities and risks in the age of digitalization and machine learning and to successfully implement innovative solutions, as shown on ite-si.de is explained. It is therefore clear that the future of condition monitoring lies in the successful symbiosis of intelligent sensor technology, advanced data analysis and the application of AI.