Extend the lifetime of your assets
Reduce the environmental footprint of your company by operating your assets longer.
Resonyx listens in to your assets, from ventilators and pumps to generators and transformers.
Non-invasive monitoring does not invalidate existing CE approvals.
Services and Solutions
Resonyx use audio analytics for a wide range of use-cases. Below, some common examples are listed.
Audio On Demand
Our simplest solution will enable you to request snippets of the live data streams from your assets at any time, augmenting the data snippets by adding statistical insights from our basic analysis models.
Analytics as a Service
By buying or renting our Resonyx edge device, our team will partner with your subject matter experts to fine tune model parameters or develop new, customised models tailored to your needs.
Abnormality Scores
By ingesting operational data, our AI solution will over time learn the normal sound of your assets, and use this insight to detect and alert operators when the sound of your machines deviate from normal.
Detections and Alarms
Resonyx is designed to identify and categorize sounds that deviate from normal. These sounds can be categorized as either anticipated, such as those generated by short-lived mechanisms within your assets, or unanticipated. In the former scenario, our solution can be utilized to monitor these mechanisms by conducting statistical analyses on their frequency and duration, or by implementing specialized sound analysis algorithms. In the case of unexpected events, our solution allows for the creation of reports and alarms for prompt notification.
We Integrate With Your Ecosystem
Recognising the uniqueness of your business, our cloud-agnostic administrator module is designed to seamlessly operate on your infrastructure, irrespective of your chosen cloud provider or private cloud configuration
Analysis Models
Our amazing data scientists have carefully developed a series of parametric models that can be tuned into the sound of almost any machine to extract, detect and classify different aspects of the emitted sound.
Feature Extraction
11 models extracting information from the sound, including loudness, spectral distribution and harmonic distortion.
Detection
3 models detecting sudden events and creeping changes.
Classification
2 models that accurately classifies specific events based on their unique audio profile.