Hannover / 04/23/2018 - 04/27/2018
Hannover Messe 2018
Noise can often be used to determine whether a moving or rotating part is functional or not. Scientists at Fraunhofer IDMT are trying to teach machines how to "listen" in order to identify defective products or defects in the production process. At this year’s Hannover Messe the combination of sound recording and analysis using machine learning methods, taking into account privacy and security aspects will be demonstrated.
Condition monitoring of axial piston pumps is done by means of intelligent acoustic sensor technology developed by Fraunhofer IDMT. To demonstrate the functional principle of the solution, typical sounds occurring during operation of the pumps are replayed via a loudspeaker. The acoustic sensors receive these audio signals and detect the sounds indicating machine malfunction. Via a wireless interface, these error-indicating signals are transmitted to a graphical user interface and visualized in order of appearance.
We improve the privacy of information from data providers by decoupling them from the real world. With a new virtual, non-traceable identity such data can still be differentiated, to be combined with information from partners and even competitors for more business, efficiency and trust.