Senstrali Triage was developed on a test rig, with used and failed rollers from WA mine sites, by an engineer with 20 years of designing instruments for extreme environments. This page shows what was built, what it found, and what it doesn't show yet.
Three steps, each forced by what the last one taught.
Used conveyor rollers on a timber stand, spun by hand and recorded as they coasted to a stop. Rough, but it showed straight away that some used rollers were clearly different from others, and what the next rig had to control.
An automated bench that runs each roller through a sweep of speed and load. The roller is driven by an elastic belt so the measurement hears the roller, not the motor. Its job is to check the sensors against rollers whose condition is known.
The lab electronics were hand-built on protoboard. The field version uses boards that can be ordered and assembled quickly, with many nodes on one cable inside conduit. Two lessons drove the change: bare cable is exposed to damage and UV, and hand-assembling each node took too long.
Each node clamps to the stringer and picks up the bearing through the frame, not through the air. It takes four measurements: ultrasonic, vibration, audio band and temperature.
A batch of failed rollers came to the rig from a WA mine site. One carried the signature of an inner race defect: impacts at the inner race defect frequency, with the sidebands you expect as the defect turns in and out of the load.
It showed up in the audio band, vibration and ultrasonic measurements taken on the frame, the same place a node sits on a conveyor.
A physics model where the defect frequencies come from impacts rather than a formula, built to match what the rig measures. Try the bearing fault simulator.
End-of-life means pulled from service, not proven failed. The rig is how we find out which is which.
Both recorded through the frame by the same node, at the same speed and load. Levels are as recorded, not normalised. The ultrasonic clips are shifted down so you can hear them.
Rig results on real used and failed rollers, and a multi-node system logging on one cable and reporting over the mobile network.
Detection on a working conveyor under load, at scale, and a false-alarm record. Enclosures are designed to meet IP67; environmental ratings are not yet tested.
Pilots on operating conveyors, with results reported as false alarms per 1,000 monitored zone-days and checked against the site's own work orders. Run-to-failure testing on the rig is planned, as an input to Senstrali Forecast.
Scope, duration and success measures agreed with your site before anything goes on a stringer. Why roller monitoring matters.
