Technology
Resonyx is built on cutting-edge acoustic analysis technology, transforming sound into actionable insights. Designed for industries where precision, safety, and reliability are important, our technology bridges the gap between raw audio data and intelligent decision-making.
A State-Of-The-Art
Acoustic Analysis Platform
By listening to and analyzing specific sound patterns, Resonyx provides real-time insights that empower proactive maintenance, enhance safety, and ensure seamless operations.
How it works
Capturing Sound
Resonyx utilizes high-quality microphones and, in some applications, complementary sensors to capture acoustic and vibrational data directly from your operations. This, together with our dynamic, containerised data pipeline, enables us to listen to critical processes and detect sounds that signify changes, anomalies, or events.
Analyzing with AI
At the heart of Resonyx is a sophisticated AI model trained to identify specific, recognizable sound patterns. Our technology doesn’t just monitor—it learns. By understanding the acoustic signatures of your machinery, processes, or environment, Resonyx can distinguish between normal operations and potential issues.
Processing Locally
Resonyx’s edge-based processing ensures that analysis happens on-site, without relying on constant internet connectivity. This allows for real-time results, even in remote or high-risk environments. Moreover, our proprietary container orchestration algorithm provides a fine-tuned data scheduling that can allow cloud control, monitoring, and storage even for corner cases where the connectivity is spotty.
Delivering Insights
When a critical sound or deviation is detected, Resonyx provides immediate feedback through automated alerts or visualized dashboards. This actionable information helps your team respond quickly, reducing downtime and improving operational reliability.
Key Innovations
Edge-Based
Intelligence
Advanced Docker runtimes scheduling, and efficient REST api connections reduce the dependency on the cloud even in those cases where it is required for remote monitoring and data acquisition. Offline buffering and smart upload through an internal database allow for remote operational data injection..
Custom Sound Recognition
Resonyx AI models detect and classify sounds specific to your environment. From the hum of a compressor to the impact of machinery, the system learns and adapts to operational sound patterns for more accurate monitoring.​
​
Built for Industrial Applications
Resonyx applies its technology across industries, from detecting wear in machinery to monitoring safety in dynamic environments. Designed for versatility, it scales to different applications and integrates with existing systems.​
​
​
AI Powered Analysis
Resonyx is designed to continuously record and analyze up to 16 audio channels at 48kHz. It features a cloud-agnostic control system utilizing the MQTT protocol, and efficient serialized REST api data connections. Communications go through our hardened Kubernetes endpoints, with state of the art cryptography and authentication flows.
​
At the edge, multiple models operate simultaneously to extract and interpret audio features. These models range from simple calculations, such as total loudness detection, to advanced machine learning algorithms capable of classifying specific audio patterns. To minimize cloud costs, only essential information—such as time series data and relevant audio snippets—is uploaded, achieving significant cost reductions while maintaining data integrity.
Technical articles
Read more about sound, audio analysis and the Resonyx technology in our articles.
The Hard Facts
Resonyx Device types
We offer several types of edge devices and packages based on the complexity and requirements of each application.