Silicon Labs Helps Waites Turn Sensor Data into Insight for AI-Powered Condition-Based Monitoring
|
Waites Provides Superior Condition Monitoring in Industrial Environments
In today’s highly competitive industrial environments, it would be a mistake to dismiss all sensor-based CbM solutions as essentially the same. This is especially true in large-scale warehouses and factories that would require miles of cable and other expensive physical infrastructure for traditional wired systems, not to mention the time needed to manually collect measurements.
Waites is showing just how much is possible with connectivity by delivering high-fidelity wireless vibration monitoring at scale in a way that traditional systems can’t match. By designing chip-down hardware, synchronized data acquisition, on-chip AI analysis, and wireless networking, Waites is giving its customers the precision and responsiveness required to keep equipment up and running.
Eliminating Downtime with AI-Powered Preventive Maintenance
Industrial facilities operate numerous motors, pumps, gearboxes, conveyors, fans, and more, and all of these have points of failure unique to them. Unscheduled downtime due to failure can have enormous financial impact, oftentimes incurring costs that far exceed the cost of the machinery. Many of today’s vibration systems only alert users when a threshold is crossed but fail to detect early-stage defects or diagnose the root causes. Factory floors and other industrial environments require decisive action at the earliest signs of equipment failures. Consistent assessment of essential infrastructure and operations can help detect, alert, and address operational irregularities. By analyzing and interpreting the status of various types of motorized moving machinery, businesses can minimize unplanned downtime, enhance asset lifecycle management and maintenance deployment, and ultimately, reduce safety risks.
However, to monitor hundreds or thousands of assets across a plant, wireless sensors need to be low-power, easy to deploy, and communicate reliably without requiring integration.
Waites has been a leader in this space for nearly two decades, offering plant and facility managers a comprehensive view of their operations through 24/7, plug-and-play online monitoring. Its system features a wide range of sensors and gateways for monitoring and analyzing tri-axial vibration and temperature data over long distances needed for facility coverage.
The Challenge
To deliver high-fidelity, synchronized, and scalable wireless monitoring to detect subtle equipment faults while avoiding costly downtime that traditional walk-around inspections often miss.
The Solution
Waites built a full-stack solution, including custom antennas, firmware, and software, based on Silicon Labs' wireless SoCs, to deliver not only superior condition-based monitoring (CbM) sensors but also provide advanced AI analysis to prescribe specific remediation guidance.
The Result
This approach makes it possible to capture more accurate data, but also pinpoint issues such as misalignment or bearing defects, rather than just generating alerts. As a result, Waites customers collectively save more than $300 million in downtime costs annually.
Why Time Synchronization Matters in Condition-Based Monitoring
A key advantage Waites has over its competitors is that it does all of the data analysis itself. With in-house vibration analysis and its AI machine learning algorithms trained on 8 trillion sensor readings from half a million sensors installed around the world, Waites can provide much more granular insight into potential issues.
This makes time synchronization an important part of the Waites solution. By aligning vibration data from multiple sensors in real-time, Waites can distinguish between similar fault signatures, such as misalignment, unbalance, or structural looseness, and determine the exact issue. The better they can analyze their data, the more sophisticated their analysis will be.
Time synchronization also enables correlation across machine trains, where a single adjustment can ripple through connected shafts, couplers, or bearings. Without synchronized measurements, many of these insights would be lost, reducing diagnostic accuracy to little more than generic alerts.
Silicon Labs MG24: Powering the Waites Data-Intensive AI Sensors
Waites turned to Silicon Labs and the EFR32MG24 multiprotocol SoC, to develop a predictive solution that scales across facilities, equipment types, and operating conditions. The reliability and secure communication of the MG24 made it an ideal fit for its data-intensive, battery-powered sensors. The peripheral reflex system (PRS) of the MG24 allows the acquisition system to capture data at higher rates while offloading time-critical tasks from the CPU. This makes it possible to run more sophisticated filters, improve anti-aliasing, and deliver higher fidelity signal analysis without sacrificing efficiency.
And because EFR32 wireless devices share a common software ecosystem and cross-device compatibility, Waites is able to take advantage of time synchronization in earlier generation solutions that feature other Silicon Labs devices. This means algorithms for vibration analysis and filtering can be reused without major rewrites, even when moving from earlier Series 1 devices to newer Series 2 parts like the MG24. The stability across platforms reduces development overhead, accelerates migration, and ensures consistent performance in condition monitoring applications.
Benefits of Using AI-Powered Systems During Condition Monitoring
Utilizing the MG24 built-in AI/ML matrix vector processor (MVP) hardware accelerator, Waites can complete sizeable complex mathematical operations of time-series data up to 8x faster while consuming 6x less energy. Its AI-powered systems also unlock timely interventions by training a model, detecting anomalies, and continuously monitoring equipment parameters. AI/ML uncovers hidden insights into sensor data, turning condition monitoring into a proactive, cost-effective, and data-driven process. Data from sensors, including vibration waveforms, temperature, and meta metrics is analyzed by on-chip AI and then streamed to Waites’ cloud system via cellular or gateway relays with no required integration to the facility’s IT network.
The advanced radio tuning available in Silicon Labs’ RAIL (Radio Abstraction Interface Layer) library provides Waites with control over wireless performance while simplifying protocol implementation. It includes advanced features like carrier sense, energy detection, collision avoidance, and exponential backoff to map network connectivity, monitor signal strengths, and precisely time network transitions for optimal reliability and efficiency.
Utilizing time synchronization, a unified developer experience, and flexible performance, Waites is able to deliver performance monitoring insights that help customers detect potential faults earlier, act with confidence, and avoid costly downtime.
With the help of Silicon Labs, Waites is demonstrating how connectivity, precise data acquisition, and AI can elevate condition monitoring beyond basic sensing.