“If you can measure it, you can manage it.” George McMurtry heard this principle early in his career, coming from an instrumentation background, and it still guides how he approaches reliability.
To make sound predictive maintenance decisions, teams need to understand how each machine is operating, and that depends on reliable vibration and temperature data.
In this DynaTalks episode, Deyvid Souza, from Dynamox, talks with George, director at Intercom and Dynamox partner in Australia, about continuous monitoring, data density and how reliability work is changing in Australian plants.
Here are the main takeaways from the conversation.
Where is your asset on the P-F curve?
With route-based collection every four to six weeks, a team can’t tell where an asset sits on the P-F curve.
A machine can move from potential failure (P) to functional failure (F) between two readings, and the breakdown comes as a surprise.
Continuous monitoring reduces that blind spot. When vibration rises, the team knows something has changed, and spectral analysis shows what changed and where the fault comes from.
George gives the example of five pumps monitored continuously. When an outer race defect appears on one of them, the team follows how it develops along the curve and chooses the right moment to intervene.
The same logic applies to lubrication and predictive maintenance: a rise in acceleration caused by lack of lubrication shows up early, so the bearing can be greased before it runs dry.
Repeatable data from the same measuring point
For George, data quality starts with repeatability. A technician measuring temperature with a laser gun rarely hits the exact same spot twice.
A wireless sensor fixed to the machine always measures from the same location, so readings can be compared over time.
This is one of the keys to getting reliable data from wireless vibration sensors.
Dynamox works with two complementary types of data:
| Data type | What it captures | How often |
|---|---|---|
| Telemetry | Acceleration, velocity and temperature on three axes | Every 1 to 10 minutes |
| Spectral data | Triaxial frequency spectrum, used for root cause analysis and fault-frequency tracking | Once a day, and up to eight or more times a day during severe failures |
Data density: from snapshots to 1,440 points a day
George compares data density to health care: it’s the difference between regular blood tests and waiting until you feel sick to see a doctor.
A snapshot every four to six weeks is no longer enough.
With telemetry sampled every minute, a single sensor generates 1,440 data points in 24 hours.
That volume feeds AI algorithms around the clock, helping teams anticipate failures with more precision, as explored in our article on wireless sensors and AI in vibration analysis.
A different 12-hour shift
Online monitoring also changes the routine. Instead of spending hours of a 12-hour shift walking the plant to collect data, analysts can start at their desk with a cup of coffee.
There, they open the Decision Making Assistant (DMA), the main entry point to the Dynamox platform.
If five red alerts arrived by email, they analyze each one, grab a bottle of water and head out to inspect only the machines that need attention.
From pumps to bucket wheel reclaimers
In Australia, George’s team monitors “anything that rotates or moves”: pumps, motors, gearboxes, conveyors, screens, crushers, SAG mills and ball mills.
These assets are spread across sectors ranging from mining and steelmaking to food production.
Sensors also support plants that want to push throughput beyond nameplate capacity.
On bucket wheel reclaimers, for example, sensors on the structure show how the frame responds as more material is reclaimed.
Fast deployment and a new generation of technicians
What won George over was how simple deployment is: plug in the gateway, power it up, and the sensors connect to the cloud platform.
On one site visit, technicians were recording exciter temperatures in spreadsheets. The data was live on the platform before the team walked out the door.
Larger projects scale quickly: around 500 sensors installed in six days, or magnetic sensor kits ready in about 15 minutes.
George also sees online monitoring as a way to move beyond fixed calendars in predictive maintenance.
A pump replaced every six weeks can last eight or twelve, based on its actual condition.
With experienced professionals retiring around the world, this kind of system gives younger teams the data confidence they need to keep plants running.
Watch the full DynaTalks episode
Want to hear the whole conversation? Watch the DynaTalks episode with George McMurtry and see how continuous data supports better maintenance decisions in the field:
If your team wants to move from route-based collection to continuous monitoring, talk to Dynamox specialists and learn more about our wireless vibration sensors.