EASY-TO-USE DIAGNOSTICS FOR INTELLIGENT MAINTENANCE DECISIONS
What Is NL Analytics?
NL Analytics lies at the heart of what we do: turn sound into value. We developed NL Analytics to carry out this function. NL Analytics is central to our intelligent solutions that solve specific problems, such as detecting compressed air leaks in factories or partial discharges in power grids.
You might encounter devices marketed for these purposes that locate various noise sources or, at best, maybe show decibel levels. We do things differently at NL Acoustics. To begin with, our comprehensive, machine learning-powered solutions detect, distinguish, and locate distinct sounds, whether it is compressed air leaks or partial discharges you are trying to find.
Additionally, our advanced NL Analytics calculates leak and size cost estimate for compressed air leaks, and assesses the severity of partial discharges, among many other useful features.
Find out more about these features in the sections below.
How Does NL Analytics Work?
NL Analytics automates problem solving, saves you time, and helps you make intelligent maintenance decisions.
Using devices without analytics requires the users to have a more in-depth understanding to interpret the results. We combine acoustic and problem-specific expertise and know-how with extensive data, advanced algorithms, and machine learning to give our users clear-cut results with minimal training required. In this way, single users do not need broad knowledge to translate the results but can instead focus on fixing the issues our solutions identify. With the assistance of NL Analytics, you can turn sound into valuable information.
NL Analytics in Our Solutions
NL Camera for Partial Discharge Detection
On-device partial discharge detection, localization, and PRPD pattern
Users have their first contact with NL Analytics when the NL Camera detects and localizes a partial discharge based on the sound it emits, recognizing and distinguishing it from other sounds. The on-device analytics then shows the PRPD (phase-resolved partial discharge) pattern that depicts the partial discharge’s behavior and helps to classify the discharge type.
The severity assessment is an unmatched tool for easily discovering whether the detected partial discharge poses a problem, how threatening it is, and what actions the user should take next to address the issue. The severity assessment is included in the two machine learning-powered NL Analytics software options, the NL Cloud and the offline version, NL Camera Viewer Pro. Both use the PRPD pattern to first categorize the detected partial discharge. There are several different types of partial discharges, some more serious than others. The most common are corona and surface discharges.
To determine the partial discharge’s severity, the NL Analytics software applies the knowledge of partial discharge types and data to similar problems. We use this data to define a scale for a partial discharge’s severity and turn it into an algorithm for NL Analytics. Now NL Analytics can evaluate how severe a detected partial discharge is on a four-tier scale that designates each discharge as being either low, medium, high or very high.
Furthermore, NL Analytics gives a detailed description of the problem as well as recommends which actions to take next to fix the issue.
For instance, a corona discharge is often low in severity. In our example (see Image 1), the NL Camera detected corona discharge on a disc insulator string of a 220 kV electrical system from a 15-meter distance. NL Analytics determined that a corona discharge of this strength is harmless. Since the partial discharge in question was not serious, NL Analytics recommended no further action so long as the discharge did not proceed to cause other problems, such as disturbing noise or power loss.
Image 1. An example of the NL Analytics softwares’ severity assessment.
NL Camera for Air Leak Detection
Users encounter NL Analytics for the first time when the NL Camera detects and localizes a compressed air leak, recognizing and distinguishing it from other sounds, even in noisy factory environments. The NL Camera measures the sound emitted from the leak, and NL Analytics calculates a real-time leak size estimate. Knowledge of the leak size also enables NL Analytics to calculate an annual cost estimate of the leak.
This information is available in the NL Cloud and NL Camera Viewer Pro, where users can inspect pictures of leaks more closely and export them into an ISO 50001-compatible report.
NL Sense for Automated Asset Monitoring
NL Analytics in the NL Sense is tailored to solve specific problems in asset monitoring. Through NL Analytics, the NL Sense identifies problems before they become critical. This allows you to react when problems arise without having to constantly audit perfectly functioning assets manually.
The NL Sense’s analytics learn the normal operating sounds of an asset it is attached to, creating a unique acoustic fingerprint for the asset by tracking it in real time. Following this, NL Analytics can recognize even the smallest deviations in the operating sounds and alert the users. Such outliers are indications of the onset of a failure chain.
The NL Sense’s analytics collect an immense amount of data that the machine-learning powered solution can then transform into a useful, easily understandable format. You can view this information either integrated into your existing asset management system or in the accompanying NL Cloud software.