ACME 4.0 – Self-adapting sensor systems for acoustic condition monitoring in Industry 4.0 applications

Networked sensors and measuring systems are increasingly used in industrial production to make manufacturing processes more reliable. So far, acoustic sensor systems have been used in only a limited number of applications. In the ACME 4.0 project, partners from industry and research will investigate the potential of acoustically monitoring machines and production processes. For this purpose, the research partners will develop methods for real-time detection of acoustic events in the high ultrasound range using structure-borne noise sensors. Using methods from computer-based acoustic event detection, the goal is to permit detection of abnormalities or faults in the production process more easily and reliably than in the past. Within the scope of the project, this technology will be tested in quality control in semiconductor production and in fault detection on axial piston pumps. The goal is to develop an energy-autonomous, highly-integrated sensor platform that can be integrated wirelessly in industrial installations.

In this project, the Fraunhofer IDMT's Oldenburg Branch for Hearing, Speech and Audio Technology HSA is further developing existing acoustic sensor systems for airborne sound so that these systems can also be used for the frequency ranges of infrasound and ultrasound. The sensor node platform will be extended by an interface for wireless data communication. As part of the project, the scientists are building an audio database and a catalogue of typical acoustic signal characteristics. The goal is to develop robust recogniser algorithms for industrial applications using machine learning methods.