Detect undesired sounds in real time automatically with Sounce
Whether in quality assurance or component qualication - noise provides information about product and process quality. However, the analysis and detection of noises is complex, often linked to additional working hours and based on the subjective perception of the engineer. Failures detected too late or not at all must be corrected with great expense and effort.
Based on a deep-learning approach, Sounce detects noise reliably and continuously. Constant testing speeds up the defect elimination process and facilitates root cause analysis. With reduced inspection costs and faster results, Sounce helps to ensure product quality at an early stage and allows the engineer to devote time to other activities. Station monitoring and fault detection is visualised in a web application. This ensures that the tested quality features are sustainable and cost-effcient.
The solution allows early and reliable fault detection in the production process as well as a reduction of manual monitoring and expensive rework costs on both end-of-line test stands as well as in-line. Furthermore, an active learning system automatically detects known as well as newly occurring defects. Sounce is capable of easily tracing defects through continuous data recording and documentation in real time. Also, remote monitoring via a cloud-based web application is possible.
- Static hardware setup as well as repetitive testing
- Training data for model training
- Medical Technology