IAV's AI and audio analysis experts have refined the identification and localization of acoustic anomalies in any technical system, offering you a complete professional service for an end-to-end audio analysis solution. The pre-processing as well as the AI training and post-processing are optimized to achieve high detection and classification rates of sound events and anomalies in acoustic or vibrational signals. The entire process chain is integrated into the powerful processing pipeline IAV SonicSeek running on AWS services. This ensures that computing power is not only optimally utilized in relation to costs, but is also instantly adaptable and scalable to your specific needs.

Key AWS services included in the solutions architecture are Amazon S3, AWS Lambda, Amazon SageMaker, Amazon EC2, Amazon EKS, Amazon DynamoDB, Amazon Route 53, Amazon CloudFront, Amazon ECR, Amazon RDS.

IAV SonicSeek consists of service components including consultation on the data acquisition setup including suitable audio equipement, data analysis, training and evaluation of use case specific AI-models, and serving these models by API or user interface. The components can be combined for an end-to-end analysis solution of sounds and noises. It is applicable to a wide range of use cases in diverse industries, including, but not limited to:

• Machine condition monitoring
• Predictive maintenance and after sales product monitoring
• Quality assurance in production
• Squeak and rattle (NVH testing)
• Surveillance

Based on our fully automated training pipeline and optimized neural networks for audio analysis you will achieve high accuracy results with short setup times. Our IAV SonicSeek Audio Analyzer tool supports you at the analysis of recorded audio signals and event detection performance. We provide the trained model as Software as a Service with easy to integrate REST API or white-label web-frontend customized for your needs.

Further information can be found on our product homepage or view our product video.

The cost of IAV SonicSeek is determined by the task complexity and the used service components. Please contact us for further details.