"Many cooks may spoil the broth, but this does not apply to the automatic condition monitoring of complex machines and systems," explains Professor Dr. Volker Lohweg, Institute Director at inIT and expert in the field of sensor and information fusion. "The inclusion of many different data sources, including not only sensors and actuators but also process data from the databases of the plant operators' ERP and MES systems, is essential in order to record the current status of modern technical systems."

In doing so, the researchers at inIT make targeted use of data that is actually superfluous - the so-called redundancies that arise due to the use of many different data sources. Redundancies are used profitably to determine and reduce conflicts in the statements of the individual data sources, for example in different machines along a production line. Conflicts are a sign that the statements of one or more data sources do not match the statements of the others. There are many reasons for this and can, for example, indicate incorrect parameterization or a defect in various sensors.

Research in the "ReWork" project
A current research project at inIT in the field of sensor and information fusion is the "ReWork" project. Focusing on the production of components for the aviation industry, the project is developing solutions to optimize production processes and, in particular, to reduce waste during repairs to aircraft wings and other components.

"The production of aircraft components is highly complex and requires maximum accuracy and precision in order to avoid material damage or breakage during flight and thus disasters," explains Lohweg. The quality control of industrially reworked components is essential, but cannot be effectively implemented with currently available techniques such as ultrasonic testing technology or X-ray inspection.

A new process for automated quality control has therefore been developed at inIT with partner Owita. Using an imaging process, sensors generate a kind of network of wave-shaped signals ("lamb waves") on the aircraft parts to be inspected and measure repair areas on the components. The measurement signals are evaluated automatically by computer-controlled algorithms that use the transmitted sensor information to assess and monitor the quality of the reworking and ultimately pass it on to the user.

"The sensor and information fusion ensures an increase in process reliability and holistically optimizes the individual steps along the process chain," says Lohweg, summarizing the benefits of the project. as