At the core of our solution is a powerful combination of artificial intelligence and heuristic algorithms, developed and refined using over one million hours of real-world data. This robust foundation enables highly accurate detection and classification of events along critical infrastructure, dramatically reducing nuisance alarms.
These advanced algorithms deliver actionable data that supports our clients’ concepts of operations, enhances operational efficiencies, and protects high-value assets. The system provides an unparalleled, uninterrupted overview of the full asset, enabling real-time decision-making and situational awareness. By optimizing signal processing for specific activities—such as detecting digging while ignoring animals—our technology ensures that only the most relevant and critical alerts are surfaced. Powered by a continually expanding library of labelled data and state-of-the-art machine learning models, our platform offers precise, reliable, and field-proven performance.
ONYX™ enables improved signal processing.
Processing splits moving and
stationary activity.
20dB (x10) improved acoustic noise performance provides better input data into detection and classification algorithms.
Accurate classification of right of way activities.
Utilising advanced signal processing techniques data is optimised for the specific activity trying to be detected e.g. digging signals.
After this AI/Machine learning algorithms are used to correctly classify the event and equally as importantly to disregard signals that could cause nuisance alarms.
This enables the the ability to correctly classify digging but not alert on animals/cattle/horses on the right of way.
Detection & Classification Algorithms
Using Machine Learning.
Detection and classification of events along an assets achieved using Machine Learning based algorithms.
Large and expanding library of labelled data use to train each iteration of ML model developed.
Backed up with advanced AI algorithms and ever advancing machine learning, all trained on cross-industry applications. Each unit is preloaded with a library of greater than 1,000,000 hours of real-life data providing the highest detection rates with a drastic reduction in nuisance alarms.
This ever-learning life-cycle ensures you get the most out of your system in your environment.
Monitoring your data in real-time, any time. The software is served from each ONYX™ Sensing Unit connected to Local or Wide Area Networks. This provides full remote access to all the functions including, engineering setup, instrument diagnostics, data visualization, data processing and alert notification.
Full software control. All the acquisition sensing parameters such as pulse width, gauge length, sample rate and decimation factor are software defined and can be quickly and easier changed to enable the optimum settings use in each deployment.
Industry leading algorithms. Based upon many decades worth of combined experience the Sintela team have developed a suite of application specific, detection and classification algorithms providing unrivalled levels of detection with low rates of nuisance alarm.
Web based user Interface. Each ONYX™ incorporates its own web server providing the user with remote web access to a modern user interface from any web browser on a secure intranet / internet connection.
Distributed architecture. ONYX™ has been designed around a fully distributed architecture eliminating single points-of-failure in multisystem networked deployments.
Automatic setup with fiber monitoring. An automated installation process simplifies the setup, commissioning and testing stages of the ONYX™ implementation.
Open architecture. ONYX™ is an open system meaning that a customer can load and operate their own software upon the ONYX™ hardware if they prefer.
By harnessing the power of quantitative data ONYX™ accurately classifies and geolocates all moving objects and potential threats within 6 metres both in real-time and historically, putting you in control to respond appropriately.
Quantitative Measurements
ONYX™ is a fully coherent, phase-based DAS system which enables the use of sophisticated event classification algorithms to achieve a low nuisance alarm rate.
Industry leading sensitivity. Capable of resolving an axial strain of 1.5 pico-strain using standard single-mode optical fiber.
Integrated compact design All the optical sensing components, electronics and processing hardware have been incorporated into one compact unit.